57DETAILED 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 . This action is responsive to the amendment filed July 6, 2026.
The status of the claims is as follows.
Claims 1, 2, 7-9, 14-16, and 21 are amended and Claims 4, 6, 11, 13, 18 and 20 are cancelled. Claims 22-24 have been added. Claims 1-3, 5, 7-10, 12, 14-17, 19, 21-24 are currently pending.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 5; 8-9, 12; 15-16, 19, 22-24 are rejected under 35 U.S.C. 103 as being unpatentable over Houdray et al. (US20130231756A1, hereinafter “Houdray”) in view of Matsuda (JP2019082783A) in view of Khafaf et al. (“Bayesian regularization of neural network to predict leakage current in a salt fog environment” [2018], hereinafter “Khafaf”).
Regarding Claim 1,
Houdray discloses A method comprising: determining, based on a first set of received resistance values, a first present insulation level for a first insulation layer on a first cable providing power to a first electronic device based on a first current leakage test (Houdray [Fig. 8];
PNG
media_image1.png
85
395
media_image1.png
Greyscale
Houdray [0045]; “The maintenance device preferably comprises at least one current leakage detector or an insulation monitor located on or connected to at least one line of the electric installation to be monitored to provide signals representative of a current earth leakage or of an insulation fault to the processing means”, wherein detection of an insulation fault through the insulation monitor to identify a location in the cable wherein voltage/current is of an abnormal level reads on determining a level of insulation throughout the cable; wherein the voltage/current implicitly read on received abnormal resistance values used throughout detection of the insulation fault and its associated level of insulation)
determining, based on a second set of received resistance values, a second present insulation level for the first insulation layer on the first cable based on a second current leakage test (Houdray [Fig. 8];
PNG
media_image1.png
85
395
media_image1.png
Greyscale
Houdray [0045]; “The maintenance device preferably comprises at least one current leakage detector or an insulation monitor located on or connected to at least one line of the electric installation to be monitored to provide signals representative of a current earth leakage or of an insulation fault to the processing means”, wherein detection of an insulation fault through the insulation monitor to identify a location in the cable wherein voltage/current is of an abnormal level reads on determining a level of insulation throughout the cable; wherein the voltage/current implicitly read on received abnormal resistance values used throughout detection of the insulation fault and its associated level of insulation; wherein the existence of multiple electric lines of multiple electric insulations across the system reads on a second present insulation level, layer, cable, electronic device, and leakage detector tests)
generating, based on the loss function, a decision framework for the first electronic device based on the first current leakage value, the actual confidence score, and the first confidence score (Houdray [0073]; “The data provided by the detectors 53 and 54 are defined as high criticality conditions in a decision tree and are designed to disable restoration of operation of the part of the installation concerned” wherein a decision tree using current leakage detector and insulation monitor data as high criticality conditions in the decision tree reads on generation of a decision framework based on current leakage values derived through the detector, wherein generating a decision tree encompassing a plurality of electronic devices defined as “high criticality” throughout the system reads on generating a decision framework for first and second electronic devices based on their respective current leakages and confidence scores
Houdray [0059]; “The storage module 3 also stores criticality criteria and data used in a decision tree” wherein the criteria used for determining if criticality conditions are considered of high importance in the decision tree reads on a confidence score considered in the generation of the decision framework)
and displaying, … in a user interface, the decision framework for the first electronic device based on the predicted conditions for the first insulation layer (Houdray [0081]; “FIG. 7 represents a diagram of presentation of the electric equipment data showing electric installation maintenance data and a decision tree. This type of diagram can appear on monitoring and diagnostic tools 57, 76, 77” wherein the decision tree displayed on monitoring and diagnostic tools reads on displaying the decision framework in a user interface)
Houdray fails to explicitly disclose but Matsuda discloses displaying, in a table in a user interface, the decision framework (Matsuda [Page 3 Line 33]; “In this manner, the classification rule table generation device 100 generates a table that represents the determination conditions of the decision tree. Thereby, the classification rule table generation device 100 causes the user to connect the factors”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray’s method of displaying the decision framework for the deployment states of first and second electronic devices in a user interface by modifying the decision framework representation format to be Matsuda’s tables. The motivation to do so lies in how “generating a table allowing a person to understand classification rules of a decision tree” (Matsuda [Abstract]).
Houdray/Matsuda fails to explicitly disclose but Khafaf discloses determining, utilizing a supervised machine learning model, a … current leakage value and a … confidence score for the … insulation layer on the … cable providing the power to the … electronic device (Khafaf [Section 1 Paragraph 3]; “The initial value of leakage current and the slope for the LC vs. time curve at each 10 minutes during the first 5 hours of salt-fog test are adopted as the input feature vector to predict the final value of the LC of EAP [6]. ANN has been used to predict the daily variation of leakage current accurately using the environmental conditions like temperature, humidity and wind velocity as input feature vector “ wherein the autoregressive neural network predicting current leakage for insulators given historical current leakage data and environmental data reads on determining a current leakage; wherein the autoregressive neural network reads on a supervised machine learning model being utilized for determination,
Khafaf [Figure 7];
PNG
media_image2.png
558
300
media_image2.png
Greyscale
Khalaf [Section 3.2 Paragraph 2]; “It can be observed that all models of ANN predicted the measured signals with similar accuracy in terms of absolute prediction error. For the first 135 minutes the absolute error for all three models is between 10–20%.” wherein the absolute prediction error reads on a confidence score associated with the predicted current leakage being determined
Khafaf [Figure 1];
PNG
media_image3.png
549
849
media_image3.png
Greyscale
Khafaf [Section 2.1]; “A chamber with a dimension of 1m×1m×1m was utilized as a fog chamber where the fog was applied using ultrasonic humidifier with a maximum flow rate of 0.3 1/m. The rate of flow can be adjusted to change the humidity level, which is measured using a humidity sensor. During the different experiments, the humidity level was adjusted to be 80% with maximum error of ±5%. The voltage was applied using 15kVA, 220V/20kV single phase transformer. The LC current was measured as a voltage drop across a 100 Ω resistor. The experimental setup is depicted in Figure 1. The experiment is done at different voltage levels (0.6–6 kV), silicone rubber (SIR) rod lengths (6 and 10 cm) and fog conductivities (10, 15 and 20 S/cm). The water conductivity is adjusted by changing the concentration of added NaCl, and is measured by a conductivity meter. The LC of four SIR rods were monitored simultaneously for five hours under the aforementioned different conditions. Total of 20 experiments have been conducted to obtain a database for 80 rod samples as per Table 1. The rods were made from removed sheds high temperature vulcanized (HTV) SIR insulators.” wherein the insulating rods for cables in the circuit read on the determinations being performed for an insulation layer on a cable providing power to an electronic device)
Houdray/Matsuda discloses a first insulation layer on the first cable providing power to the first electronic device as well as a second insulation layer on the second cable providing power to the second electronic device. Houdray/Matsuda does not disclose determining, utilizing a supervised machine learning model, a current leakage and a confidence score. However, Khafaf discloses determining, utilizing a supervised machine learning model, a current leakage and a confidence score. By performing Khafaf’s determining of a current leakage and a confidence score upon both of the insulation layers, cables, and electronic devices of Houdray/Matsuda, the combination of Houdray/Matsuda/Khafaf thus discloses determining, utilizing a supervised machine learning model, a first current leakage and a first confidence score for the first insulation layer on the first cable providing the power to the first electronic device.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda’s method of determining current leakage in lines of an electric installation and generating a decision framework to address the current leakage to use a supervised machine learning model derived through flashover-related environmental data variables for determining the current leakage and an associated confidence score. The motivation to use predicted current leakage values over real-time current leakage values lies in how “Predicting the LC different components before they reach critical values that leads to flashover can be more helpful than predicting the flashover as this prediction can practically trigger a warning signal to the electric utility” (Khafaf [Section 1 Paragraph 7]).
The combination of Houdray/Matsuda/Khafaf discloses wherein the decision framework includes … conditions for the first insulation layer; (Houdray [0031]; “Preferably, the maintenance method comprises evaluation of a criticality level by a decision tree comprising: monitoring of the causes of tripping of an electric equipment unit, monitoring of external commands, monitoring of the events history, selectively monitoring, and/or monitoring of ageing data”
Houdray [0073]; “The data provided by the detectors 53 and 54 are defined as high criticality conditions in a decision tree and are designed to disable restoration of operation of the part of the installation concerned” wherein the high criticality evaluations by the decision tree for an electric line and its associated electric equipment unit reads on the decision framework including current conditions for insulation layers of the system). The combination does not explicitly disclose but Khafaf further discloses predicted conditions for the first insulation layer (Khafaf [Section I Column 2 Paragraph 3]; “the contamination levels can serve as a warning system and hence help in scheduling washing and maintenance routines. Both the average and maximum values of LC have been used to predict the equivalent salt deposit density (ESDD) for ceramic insulators during clean fog test using ANN [11]. In another study, peak impulse current, maximum low-frequency current, humidity and voltage have been used as an input feature to predict the ESDD. A maximum error of 18% between the actual and predicted value has been reported [12]. A similar approach has been used to predict ESDD on non-ceramic insulators during salt fog test [13]. Three different classes of ESDD level have predicted with 78% accuracy using K-nearest neighbor classifier (KNN).)”)
Houdray/Matsuda/Khafaf discloses wherein the decision framework includes … conditions for the first insulation layer. Houdray/Matsuda/Khafaf does not explicitly disclose predicted conditions for the first insulation layer. However, Khafaf further discloses predicted conditions for the first insulation layer. By using Houdray/Matsuda/Khafaf’s decision framework with the predicted conditions of Khafaf instead of the current predictions of Houdray/Matsuda/Khafaf, the combination of Houdray/Matsuda/Khafaf thus further discloses wherein the decision framework includes predicted conditions for the first insulation layer.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf’s method of determining current leakage in lines of an electric installation and generating a decision framework to have the decision framework be associated with Khafaf’s predicted conditions instead of current conditions. The motivation to use predicted conditions lies in how “Predicting the LC different components before they reach critical values that leads to flashover can be more helpful than predicting the flashover as this prediction can practically trigger a warning signal to the electric utility” (Khafaf [Section 1 Paragraph 7]).
The combination of Houdray/Matsuda/Khafaf discloses determination of a decision framework based on at least a first current value. The combination fails to explicitly disclose but Sun discloses creating, by the supervised machine learning model, a loss function utilizing a gradient descent between the first confidence score for the first … value and an actual confidence score, wherein the gradient descent reduces a residual error between the first confidence score and the actual confidence score (Sun [Page 10 Paragraph 3 Line 5]; “The initial SPPM may be a machine-learned regression model. Exemplary machine learning regression models may include decision tree models, deep learning neural network models, regression tree models, random forest models, and the like, or combinations thereof. Exemplary deep learning neural network models can include convolutional neural network models (CNNs), multi-layer perceptron neural networks (MLPs), full convolutional neural network (FCN) models, generative countermeasure networks (GANs), Back Propagation (BP) machine learning models, Radial Basis Function (RBF) machine learning models, Deep Belief Networks (DBNs), elman machine learning models, and the like, or combinations thereof.”
Sun [Page 10 Paragraph 4 Line 4]; “For example, the first predictive model determination module 508 may generate the target SPPM using a machine learning algorithm and training an initial model (also referred to as initial SPPM in this disclosure) based on at least two data pairs. Exemplary machine learning algorithms may include a gradient descent algorithm, a newton algorithm, a quasi-newton algorithm, a levenberg-marquardt algorithm, a conjugate gradient algorithm, or the like, or combinations thereof. Each of the at least two data pairs may include a first resolution sample image (e.g., LR data for an LR sample image) and a second resolution grid surface corresponding to the first resolution sample image. For example, the first resolution sample image may be an LR sample image and the second resolution grid surface may be an HR grid surface. The at least two data pairs may be used to train an initial model. During training, the first predictive model determination module 508 may minimize a loss function of the initial model including the updated parameters to iteratively update the parameters of the initial model. When the termination condition is satisfied, the first predictive model determination module 508 may terminate the iterative training process.”
It would have been obvious to modify the combination of Houdray/Matsuda/Khafaf to incorporate Sun’s loss function determined through gradient descent reducing a residual loss in the creation of Houdray/Matsuda/Khafaf’s decision tree decision framework. One would have been motivated to incorporate such a loss function because “Convergence may be considered to occur if the change in the value of the loss function in two or more successive iterations is less than a certain value (or approximately a constant value)” (Sun [Page 8 Paragraph 5 Lines 17-18]) and thus identify when the decision framework has reached an optimal state.
Regarding Claim 2,
The combination of Houdray/Matsuda/Khafaf teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). The combination fails to explicitly disclose but Khafaf discloses calculating a … present leakage value based on a … difference between the … current leakage value and a … prior maintenance leakage test value for the … cable (Khafaf [Section 1 Paragraph 3]; “The initial value of leakage current and the slope for the LC vs. time curve at each 10 minutes during the first 5 hours of salt-fog test are adopted as the input feature vector to predict the final value of the LC of EAP” wherein the slope of the leakage current graph created through the difference between two leakage values over some time t implicitly reads on calculating a difference between a current leakage value and a leakage value at another time before the current leakage value (thereby prior to the maintenance), thus reading on the calculation of a present leakage value)
and predicting the … current leakage and the … confidence score based on … electronic device data, … weather data, … maintenance data, and the … present leakage value (Khafaf [Table 2];
PNG
media_image4.png
570
904
media_image4.png
Greyscale
Khafaf [Section 2.2 Paragraph 9]; ”The training datasets for both the fundamental and the third harmonics were selected. Table 2 depicts samples of the data used for training the neural network along with the experimental conditions.” wherein the training dataset sample constitutes electronic device data (present voltage, length of cable insulation rods) in which the prediction is trained and thereby based on;
Khafaf [Section 1 Paragraph 3]; “ANN has been used to predict the daily variation of leakage current accurately using the environmental conditions like temperature, humidity and wind velocity as input feature vector” wherein weather data is used in the prediction;
Khafaf [Table 1];
PNG
media_image5.png
413
935
media_image5.png
Greyscale
Khafaf [Section 2.1 Paragraph 1]; ”The experimental setup is depicted in Figure 1. The experiment is done at different voltage levels (0.6–6 kV), silicone rubber (SIR) rod lengths (6 and 10 cm) and fog conductivities (10, 15 and 20 S/cm). The water conductivity is adjusted by changing the concentration of added NaCl, and is measured by a conductivity meter. The LC of four SIR rods were monitored simultaneously for five hours under the aforementioned different conditions. Total of 20 experiments have been conducted to obtain a database for 80 rod samples as per Table 1” wherein the training dataset of Table 2 is created through the maintenance data described in Table 1 (test conditions maintained during experiment are recorded in applied voltage, No. of conducted tests) by which the training dataset of Table 2 was derived under. Thus, the training data is based on maintenance data and thereby the current leakage prediction is based on maintenance data;
Khafaf [Section 3.2 Paragraph 1]; “The trained ANN takes 5–7 past values of LC, depending on the ANN model used, as an input to predict the next future value of the LC and the process is repeated until prediction is stopped” wherein the past values of LC implicitly calculating a present leakage value during gradient descent (wherein the gradients between past leakages read on a difference) of the ANN reads on prediction of a current leakage based on a present leakage value)
Houdray/Matsuda discloses a first difference between a first current leakage value and a first prior maintenance leakage test value for the first cable as well as a second difference between a second current leakage value and a second prior maintenance leakage test value for the second cable. Houdray/Matsuda does not disclose calculating a present leakage value. However, Khafaf discloses calculating a present leakage value. By performing Khafaf’s calculating of a present leakage upon both of the first and second cables’ calculated differences of Houdray/Matsuda, the combination of Houdray/Matsuda/Khafaf thus discloses calculating a first present leakage value based on a first difference between the first current leakage value and a first prior maintenance leakage test value for the first cable; calculating a second present leakage value based on a second difference between a second current leakage value and a second prior maintenance leakage test value for the first cable.
Houdray/Matsuda discloses first electronic device data, first weather data, first maintenance data as well as second electronic device data, second weather data, second maintenance data. Houdray/Matsuda does not disclose predicting the current leakage value and the confidence score. However, Khafaf discloses predicting the current leakage value and the confidence score. By performing Khafaf’s predicting of a current leakage and a confidence score upon both of the electronic device data, weather data, and maintenance data of Houdray/Matsuda, the combination of Houdray/Matsuda/Khafaf thus discloses predicting the first current leakage value and the first confidence score based on first electronic device data, first weather data, first maintenance data … ; and predicting the second current leakage value and the second confidence score based on second electronic device data, second weather data, second maintenance data … . Since the combination of Houdray/Matsuda/Khafaf already discloses the first present leakage value as well as the second present leakage value, the combination of Houdray/Matsuda/Khafaf thus further discloses predicting the first current leakage value and the first confidence score based on first electronic device data, first weather data, first maintenance data, and the first present leakage value; and predicting the second current leakage value and the second confidence score based on second electronic device data, second weather data, second maintenance data, and the second present leakage value.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf’s method of predicting current leakage and associated confidence scores using a supervised machine learning model and generating a decision framework to address the predicted current leakage to conduct predictions using data based on weather as well as comparisons with previous leakage data. The motivation to incorporate previous current leakage data across different environmental circumstances lies in how ““insulators are susceptible to contamination in different forms including fog, dirt, and other environmental contaminants … Knowledge of the contamination levels can serve as a warning system and hence help in scheduling washing and maintenance routines” (Khafaf [Section 1 Paragraph 1]).
Regarding Claim 5,
The combination of Houdray/Matsuda/Khafaf teaches the method of Claim 2 (and thus the rejection of Claim 2 is incorporated). The combination already discloses sending a notification to a client device that the first insulation layer is approaching a hazardous condition (Houdray [Fig. 9];
PNG
media_image6.png
228
411
media_image6.png
Greyscale
Houdray [0107]; “Then a step 135 evaluates a criticality level by the decision tree method. A step 136 monitors the criticality level. If the criticality level is high, a step 137 disables closing of an electric equipment unit or supply of a part of the installation concerned. Then a step 138 enables acknowledgement of critical faults before local or remote reclosing of the electric equipment unit.” wherein acknowledgement of critical faults reads on a notification sent to a client indicative of hazardous conditions.)
Regarding Claim 22,
The combination of Houdray/Matsuda/Khafaf teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). The combination already discloses receiving, from an insulation tester, the first set of received resistance values over a first specified period of time and the second set of received resistance values from a second specified period of time (Houdray [Fig. 8];
PNG
media_image1.png
85
395
media_image1.png
Greyscale
Houdray [0045]; “The maintenance device preferably comprises at least one current leakage detector or an insulation monitor located on or connected to at least one line of the electric installation to be monitored to provide signals representative of a current earth leakage or of an insulation fault to the processing means”, wherein detection of an insulation fault through the insulation monitor to identify a location and occurrence in the cable wherein voltage/current is of an abnormal level reads on determining a level of insulation throughout the cable; wherein the voltage/current implicitly read on received abnormal resistance values used throughout detection of the insulation fault and its associated level of insulation; wherein the existence of multiple electric lines of multiple electric insulations across the system reads on a second present insulation level and thus the existence of an associated second set of received resistance values
Houdray [0039]; “Preferably, said processing means comprise:
means for storing data representative of an electric installation to be monitored,
means for storing data representative of settings and parameters of electric equipment units,
means for storing time-stamped data representative of events occurring in said electric installation to be monitored so as to constitute an events history, and
communication means” wherein the events history being represented through time-stamps is indicative of the monitored voltage/current/resistance values observed over some plurality period of time-stamps)
Claims 8-9, 12, 23 recite a computer program product comprising a computer readable medium with instructions to perform the methods of Claims 1-2, 5, 22. Thus, Claims 8-9, 12, 23 are rejected for reasons set forth in the rejection of Claims 1-2, 5, 22.
Claims 15-16, 19, 24 recite a computer system comprising one or more processors, computer media, and other generic equipment to perform the methods of Claims 1-2, 5, 22. Thus, Claims 15-16, 19, 24 are rejected for reasons set forth in the rejection of Claims 1-2, 5, 22.
Claims 3, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Houdray et al. (US20130231756A1, hereinafter “Houdray”) in view of Matsuda (JP2019082783A) in view of Khafaf et al. (“Bayesian regularization of neural network to predict leakage current in a salt fog environment” [2018], hereinafter “Khafaf”), further in view of Kim (US8624611B2), further in view of Turabee et al. (“Predicting Insulation Resistance of Enamelled Wire using Neural Network and Curve Fit Methods Under Thermal Aging” [2020], hereinafter “Turabee”).
Regarding Claim 3,
The combination of Houdray/Matsuda/Khafaf teaches the method of Claim 1 (and thus the rejection of Claim 1 is incorporated). The combination fails to explicitly disclose but Kim discloses receiving, from an insulation tester, a … plurality of resistance values over a … period of time for the … insulation layer (Kim [Col. 6 Line 38]; “On the basis of voltage information from different locations of isolation, isolation resistance values in the processing of voltage information can be determined”
Kim [Col. 6 Line 26]; “An exemplary embodiment of the disclosure comprises means 127 for measuring voltage of the measurement resistor for forming voltage information” wherein the measured voltage information read on as resistance values is received from a measurement resistor read on receiving the resistance values from an insulation tester)
Houdray/Matsuda/Khafaf discloses a first insulation layer as well as a second insulation layer. Houdray/Matsuda/Khafaf does not disclose receiving, from an insulation tester, a plurality of resistance values over a period of time for the insulation layer. However, Kim discloses receiving, from an insulation tester, a plurality of resistance values over a period of time for the insulation layer. By performing Kim’s receiving of resistance values over a period of time for the insulation layers of Houdray/Matsuda/Khafaf, the combination of Houdray/Matsuda/Khafaf/Kim thus discloses receiving, from an insulation tester, a first plurality of resistance values over a first period of time for the first insulation layer and a second plurality of resistance values over a second period of time for the first insulation layer.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf’s method of predicting current leakage and associated confidence scores using a supervised machine learning model and generating a decision framework to use Kim’s resistance values as data for the machine learning model. The motivation to use resistance values for model input lies in “said values [resistance values] reporting that leakage currents exists in said locations” (Kim [Col. 6 Line 40]).
The combination of Houdray/Matsuda/Khafaf/Kim fails to explicitly disclose but Turabee discloses wherein the … plurality of resistance values … represent an initial data set for the supervised machine learning model (Turabee [Section 4A Paragraph 1]; “The training phase involves data learning, by presenting it to the input layer and adjusting the various parameters of the NN to achieve the desired output value. As mentioned earlier, BP was used to train the given dataset provided (Xi, Xi+1) where Xi represents the value of diagnostic parameter at aging time ti and Xi+1 represents its prediction at aging time ti+1. The insulation resistance, along with the aging time was presented to the NN as input. “ wherein the insulation resistances presented as the various parameters input into the supervised NN reads on a plurality of resistance values representing an initial data set for input into a supervised machine learning model)
Houdray/Matsuda/Khafaf/Kim discloses both of the first plurality of resistance values and the second plurality of resistance values. Houdray/Matsuda/Khafaf/Kim does not disclose wherein the … plurality of resistance values … represent an initial data set for the supervised machine learning model. However, Turabee discloses wherein the … plurality of resistance values … represent an initial data set for the supervised machine learning model. By modifying Houdray/Matsuda/Khafaf/Kim’s calculated first and second plurality of resistance values to be Turabee’s initial data sets for a supervised machine learning model, the combination of Houdray/Matsuda/Khafaf/Kim/Turabee thus discloses wherein the first plurality of resistance values and the second plurality of resistance values represent an initial data set for the supervised machine learning model.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf/Kim’s method of predicting current leakage and associated confidence scores using a supervised machine learning model trained on resistance values and generating a decision framework to use Turabee’s resistance values to train a machine learning model. One would’ve been motivated to put the plurality of resistance values received from the combination of Houdray/Matsuda/Khafaf/Kim into a supervised machine learning model because “A requirement in safety critical applications is the early detection of insulation deterioration, which can be evaluated through a diagnostic property (e.g. insulation capacitance or resistance) of the insulation” (Turabee [Section 1 Paragraph 1]) thus allowing the model to be specialized for predicting insulation deterioration using the diagnostic property of evaluated resistances
Claim 10 recites a computer program product comprising a computer readable medium with instructions to perform the method of Claim 3. Thus, Claim 10 is rejected for reasons set forth in the rejection of Claim 3.
Claim 17 recites a computer system comprising one or more processors, computer media, and other generic equipment to perform the methods of Claims 3. Thus, Claim 17 is rejected for reasons set forth in the rejection of Claim 3.
Claims 7, 14 and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Houdray et al. (US20130231756A1, hereinafter “Houdray”) in view of Matsuda (JP2019082783A) in view of Khafaf et al. (“Bayesian regularization of neural network to predict leakage current in a salt fog environment” [2018], hereinafter “Khafaf”), further in view of Miron et al. (US8624611B2, hereinafter “Miron”), further in view of Guo et al. (CN212942208U, hereinafter “Guo”), further in view of Du et al. (CN111754465A, hereinafter “Du”), and further in view of Pakr (KR20220043548A, hereinafter “Pakr”).
Regarding Claim 7,
The combination of Houdray/Matsuda/Khafaf teaches the method of Claim 2 (and thus the rejection of Claim 2 is incorporated). The combination further discloses sending a notification to a client device providing a recommendation to perform an action (Houdray [Fig. 9];
PNG
media_image7.png
231
416
media_image7.png
Greyscale
Houdray [0107]; “Then a step 135 evaluates a criticality level by the decision tree method. A step 136 monitors the criticality level. If the criticality level is high, a step 137 disables closing of an electric equipment unit or supply of a part of the installation concerned. Then a step 138 enables acknowledgement of critical faults before local or remote reclosing of the electric equipment unit. A step 139 thus enables restoration of operation of the installation in manual, assisted or automatic manner. This step 139 can also comprise remote control of closing of an electric equipment unit or supply of an electric line. A step 140 performs communication of data to a supervisor.” wherein the notification in the form of acknowledgement of the criticality accompanied with the assisted restoration of operation reads on providing a recommendation to perform an action)
The combination of Houdray/Matsuda/Khafaf fails to explicitly disclose but Miron discloses an action selected from the group consisting of cleaning the first insulation layer of debris (Miron [0047]; “It is appreciated that insulator dusting device 36 is mounted on electric cable 37, derives electric power from electric cable 37, and cleans insulator 38 that is connected to electric cable 37”)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf’s method of predicting current leakage and associated confidence scores using a supervised machine learning model trained on resistance values and generating a decision framework to address the predicted current leakage by cleaning the insulator. One would’ve been motivated to do so because “The electric grid is known to collect dust, which, when wet by rain or humidity, may affect the conductivity or resistance of electric cables and insulator. There is therefore a need to regularly clean electric cable and insulators” (Miron [0002])
The combination of Houdray/Matsuda/Khafaf/Miron fails to explicitly disclose but Guo discloses dehumidifying the first insulation layer (Guo [Background]; “the power supply equipment mainly comprises power transmission lines, mutual inductors, contactors and the like with various voltage grades, in the work of the power equipment, the environmental humidity is an important factor influencing the normal work of the power equipment … in order to reduce the potential safety hazard of the operation process of the equipment, the service life of the equipment is prolonged … [and] the existing dehumidifying device is relatively perfect in dehumidification”)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf/Miron’s method of predicting current leakage and associated confidence scores using a supervised machine learning model trained on resistance values and generating a decision framework to address the predicted current leakage by dehumidifying the insulation equipment. One would’ve been motivated to do so because “ the environmental humidity is an important factor influencing the normal work of the power equipment, the air humidity is too high, the mildew breeding is accelerated due to the condensed moisture on the surface of the equipment, the electrical insulation strength is reduced” (Guo [Background]) thus incentivizing one to reduce the humidity to avoid such issues.
The combination of Houdray/Matsuda/Khafaf/Miron/Guo fails to explicitly disclose but Du discloses visually inspecting the first insulation layer for visible damage (Du [Page 3 Paragraph 1]; “to conduct regular inspections on the insulators … UAV inspection uses UAV aerial photography to obtain images of equipment to be inspected, and then uses image processing technology to detect equipment failures” wherein image processing of the insulator for a device failure reads on visual inspection of the insulation layer for damage)
visually inspecting an environment at a location for the first electronic device (Du [Page 3, Bottom Paragraph]; “The invention obtains the skeleton image by acquiring the binary image of the captured image, and detects the straight line segment in the skeleton image through a straight line detection algorithm to intercept a plurality of suspected insulator images; and then extracts the insulator images from the suspected insulator images according to the autocorrelation coefficient, The positioning of the insulator is realized, the distance between the sheds of the insulator and the direction of the main axis of the insulator are obtained at the same time, and the detection result of the insulator string drop is obtained. The invention can perform insulator positioning and string drop detection according to the periodic characteristics of the insulator strings and the strong anti-interference characteristics of the autocorrelation algorithm, solve the problem of poor anti-interference performance of the existing insulator positioning and string drop detection algorithms, and can effectively solve complex problems. Accurately detect the dropout of insulators in the background environment” wherein obtaining and analyzing the insulator position within a background environment within an image reads on visually inspecting an environment at a location for the electronic device)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf/Miron/Guo’s method of predicting current leakage and associated confidence scores using a supervised machine learning model trained on resistance values and generating a decision framework to address the predicted current leakage by visually inspecting the insulation layer and its environment. One would’ve been motivated to do so because “it is necessary to conduct regular inspections on the insulators to eliminate faulty insulators in time to ensure a stable power supply.” (Du [Page 3 Paragraph 1]) wherein a visual inspection is a timelier solution for ensuring stable power supply.
The combination of Houdray/Matsuda/Khafaf/Miron/Guo/Du fails to explicitly disclose but Pakr discloses confirming degradation of the first insulation layer via results for the supervised machine learning model (Pakr [Page 3 Paragraph 12]; “Based on the data received from the sensor data, a partial discharge type of the sensor data is classified using a convolutional neural network (CNN). Then, the model is trained and the pattern is identified using the image accumulated for 1 minute of the PRPD image.”
Pakr [Page 6 Paragraph 2]; “As shown, the step of determining whether the partial discharge (S240) includes the steps of collecting a pulse sequence for the signals measured in the actual use situation (S242); converting the pulse sequence into a feature vector stream (S244); and comparing with the model to determine whether it is normal (S246)” wherein the pulse sequence of the cable and its insulator being compared to that of a CNN’s output reads on confirming degradation of the insulation layer via results of a supervised machine learning model)
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify Houdray/Matsuda/Khafaf/Miron/Guo/Du’s method of predicting current leakage and associated confidence scores using a supervised machine learning model trained on resistance values and generating a decision framework to confirm that the insulation layer is degraded via the machine learning output. One would’ve been motivated to do so because “by diagnosing the occurrence of partial discharge at an early stage and taking appropriate measures, it is essential to monitor the degree of deterioration of the power cable insulator and to diagnose the partial discharge to prevent failure of power equipment.” (Pakr [Page 3 Paragraph 7])
Claim 14 recites a computer program product comprising a computer readable medium with instructions to perform the method of Claim 7. Thus, Claim 14 is rejected for reasons set forth in the rejection of Claim 7.
Claim 21 recites a computer system comprising one or more processors, computer media, and other generic equipment to perform the method of Claim 7. Thus, Claim 21 is rejected for reasons set forth in the rejection of Claim 7.
Response to Arguments
The Examiner acknowledges the Applicant’s amendments in which Claims 1, 2, 7-9, 14-16, and 21 are amended.
Applicant’s arguments on page 12 of the Remarks, filed July 6th, 2026, traversing the rejection of claims 1-3, 5, 7-10, 12, 14-17, 19 and 21 under 35 U.S.C. § 101 have been fully considered and are fully persuasive.
Applicant’s arguments regarding the 35 U.S.C. § 103 rejection of claims 1-3, 5, 7-10, 12, 14-17, 19 and 21 of the previous office action have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 JONATHAN J KIM whose telephone number is (571)272-0523. The examiner can normally be reached 11-8.
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, Matt Ell can be reached on (571) 270-3264. 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.
/JONATHAN J KIM/Examiner, Art Unit 2141 /MATTHEW ELL/Supervisory Patent Examiner, Art Unit 2141
Prosecution Insights