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 .
Priority
Acknowledgment is made of applicant’s claim for priority. The certified copy has been filed in parent Application No. PCT/JP2022/046667, filed on 12/19/2022.
Information Disclosure Statement
The information disclosure statements (IDS) were submitted on 05/30/2024, 03/06/2025, and 06/01/2026. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are: time-series signal clipping processing unit (Claim 1), normal vector registration processing unit (Claim 1), normal vector distribution density leveling unit (Claim 1-3), normal vector selection unit (Claim 4-5), abnormality determination unit (claim 6-7).
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof. The time-series signal clipping processing unit (Claim 1), normal vector registration processing unit (Claim 1), normal vector distribution density leveling unit (Claim 1-3), normal vector selection unit (Claim 4-5) are disclosed in Figure 1 as being part of the control unit, which the specification detains in [0035] that the control unit is a processing device including a CPU, thus these units are interpreted as a computer processing unit. The specification details in [0016] that the facility abnormality monitoring system includes an abnormality determination unit, with Figure 1 showing that as part of control unit 35, which [0044] of the specification details as a processing device such as a CPU, thus the abnormality determination unit is interpreted as a computer processing unit.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 2 and 3 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 details the limitation “optimize a division number of the vector space of the normal vectors such that a total number of the selected normal vectors becomes a predesignated number.
Claim 2 details the limitation “optimize sizes of the cells such that a total number of the selected normal vectors becomes a predesignated number”
Claim 3 details the limitation “optimize the sizes of the spheres such that the total number of the selected normal vectors becomes a predesignated number.”
Claim 2 and Claim 3 are dependent on Claim 1.
As Claim 1 details the limitation of “a total number of the selected normal vectors becomes a predesignated number” and Claim 2 details the exact same limitation of “a total number of the selected normal vectors becomes a predesignated number”, it is not clear whether the “total number of the selected normal vectors” of Claim 2 is the same total number as that of Claim 1 or if it is a different total number.
Furthermore, Claim 1 details “a predesignated number”, with Claims 2 and 3 also claiming “a predesignated number”. Thus it is not clear whether the “predesignated number” of Claims 2 and 3 are the same or different from the “predesignated number” of Claim 1.
Examiner interprets the limitations such that they are detailing the same limitations as Claim 1, and thus read as “the total number of the selected normal vectors becomes the predesignated number”.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-8 are rejected under 35 U.S.C. 101. The claimed invention is directed to the abstract concept of performing abstract steps without significantly more. The claim(s) recite(s) the following abstract concepts in BOLD of
1. (Original) A normal vector registration device comprising:
a time-series signal clipping processing unit configured to clip a time-series signal in a predetermined period from M (> 2) or more types of time-series signals indicating an operation state of a facility during normal operation of the facility;
a normal vector registration processing unit configured to generate an M-dimensional vector including M types of variables at a same time point from the time-series signal clipped by the time-series signal clipping processing unit, and
register the generated M-dimensional vector at each time point in a database as a normal vector; and
a normal vector distribution density leveling unit configured to divide a vector space of the normal vectors, select a predetermined number of normal vectors from the normal vectors included in each divided space so as to level a distribution density of the normal vectors, and
optimize a division number of the vector space of the normal vectors such that a total number of the selected normal vectors becomes a predesignated number.
Under step 1 of the eligibility analysis, we determine whether the claims are to a statutory category by considering whether the claimed subject matter falls within the four statutory categories of patentable subject matter identified by 35 U.S.C. 101: process, machine, manufacture, or composition of matter. The above claims are considered to be in a statutory category.
Under Step 2A, Prong One, we consider whether the claim recites a judicial exception (abstract idea). In the above claim, the highlighted portion constitutes an abstract idea because, under a broadest reasonable interpretation, it recites limitation the fall into/recite abstract idea exceptions. Specifically, under the 2019 Revised Patent Subject Matter Eligibility Guidance, it falls into the grouping of subject matter that, when recited as such in a claim limitation, covers performing mathematics or mental steps.
Next, under Step 2A, Prong Two, we consider whether the claim that recites a judicial exception is integrated into a practical application. In this step, we evaluate whether the claim recites additional elements that integrate the exception into a practical application of that exception.
This judicial exception is not integrated into a practical application because there is no improvement to another technology or technical field; improvements to the functioning of the computer itself; a particular machine; effecting a transformation or reduction of a particular article to a different state or thing. Examiner notes that since the claimed methods and system are not tied to a particular machine or apparatus, they do not represent an improvement to another technology or technical field. Similarly there are no other meaningful limitations linking the use to a particular technological environment. Finally, there is nothing in the claims that indicates an improvement to the functioning of the computer itself or transform a particular article to a new state.
Finally, under Step 2B, we consider whether the additional elements are sufficient to amount to significantly more than the abstract idea. Claim 1 does not include additional elements that are sufficient to amount to significantly more than the judicial exception because a time-series signal clipping process to clip time-series signals indicating an operation state of a facility is considered to be necessary data gathering. As recited in MPEP section 2106.05(g), necessary data gathering (i.e. acquiring a detection value) is considered extra solution activity in light of Mayo, 566 U.S. at 79, 101 USPQ2d at 1968; OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015). The additional limitation of registering the generated M-dimensional vector at each time point in a database as a normal vector is considered an insignificant extra-solution activity. Storing data in a database is considered to be an insignificant extra-solution activity because it is well known and established in the art to store data in a database. This is evidenced by Shibuya (JP2014032455) in [0008], [0013], and [0052] and Hara (JP2020027342A) in [0024]. The time-series signal clipping processing unit (Claim 1), normal vector registration processing unit (Claim 1), normal vector distribution density leveling unit (Claim 1-3), normal vector selection unit (Claim 4-5), abnormality determination unit (claim 6-7) are interpreted under 35 U.S.C. 112(f) as being computer processing units. Computer processing units are interpreted under broadest reasonable interpretation to be a generic computer element. Generic computer elements are not considered significantly more than the abstract idea and do not integrate the abstract idea into a practical application. As recited in the MPEP, 2106.05(b), merely adding a generic computer, generic computer components, or a programmed computer to perform generic computer functions does not automatically overcome an eligibility rejection. Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 134 S. Ct. 2347, 2359-60, 110 USPQ2d 1976, 1984 (2014). See also OIP Techs. v. Amazon.com, 788 F.3d 1359, 1364, 115 USPQ2d 1090, 1093-94.
Claims 2-8 further limit the abstract ideas and extra solution activities as detailed above without integrating the abstract concept into a practical application or including additional limitations that can be considered significantly more than the abstract idea.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-4, 6, and 8 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Shibuya (JP2014032455).
In regards to Claim 1, Shibuya teaches “a time-series signal clipping processing unit configured to clip a time-series signal in a predetermined period from M (> 2) or more types of time-series signals indicating an operation state of a facility during normal operation of the facility (Therefore, multiple sensors are installed on the target equipment or plant, and it is determined whether they are functioning normally or abnormally according to the monitoring criteria for each sensor. – [0004]; To solve the above problems, the present invention provides a method for monitoring the state of equipment or device by extracting feature vectors from sensor signals output from multiple sensors attached to the equipment or device, accumulating the center of each cluster and the feature vectors belonging to each cluster obtained by clustering these extracted feature vectors as training data, extracting feature vectors from new sensor signals output from multiple sensors attached to the equipment or device, selecting a cluster corresponding to the feature vector extracted from this new sensor signal from the clusters accumulated as training data, selecting a predetermined number of feature vectors from the feature vectors belonging to the selected cluster according to the feature vector extracted from the new sensor signal, creating a normal model using this selected predetermined number of feature vectors, calculating an abnormality measure based on the newly observed feature vectors and the created normal model, and determining whether the state of the equipment or device is abnormal or normal based on the calculated abnormality measure – [0010]; The equipment 101 subject to condition monitoring includes equipment and plants such as gas turbines and steam turbines. Equipment 101 outputs a sensor signal 102 that indicates its state. The sensor signal 102 is stored in the sensor signal storage unit 103. Figure 2 shows an example of listing the sensor signals 102 and representing them in a table format. The sensor signal 102 is a multidimensional time-series signal acquired at regular intervals. The table listing it consists of a date and time column 201 and data columns 202 for multiple sensor values provided on the equipment 101, as shown in Figure 2. The number of sensors can range from hundreds to thousands, and they may include sensors for the temperature of cylinders, oil, and coolant, the pressure of oil and coolant, the rotational speed of the shaft, room temperature, and operating time. In addition to representing outputs and states, control signals can also be used to control something to a specific value - [0019]; As a minimum measure, it is necessary to remove sensor signals with very low variance and sensor signals that are monotonically increasing – [0023]; Additionally, it's possible to remove invalid signals using correlation analysis. This method involves performing correlation analysis on multidimensional time-series signals. If there are multiple signals with a correlation value close to 1, indicating a high degree of similarity, these signals are considered redundant. The redundant signals are then removed, leaving only the non-redundant ones - [0024]);
a normal vector registration processing unit configured to generate an M-dimensional vector including M types of variables at a same time point from the time-series signal clipped by the time-series signal clipping processing unit (First, the processing flow in the feature vector extraction unit 104, the clustering unit 105, and the training data storage unit 106 will be explained using Figure 3. First, the feature vector extraction unit 104 receives sensor signals 102 from the sensor signal storage unit 103 for a period specified as the learning period (S301), canonizes each sensor signal (S302), and then extracts feature vectors (S303). - [0021]; In step S303, feature vector extraction is performed for each time step. One approach is to simply list the canonicalized sensor signals, but it is also possible to create windows of ±1, ±2, ... for a given time point and extract features representing the time evolution of the data using a feature vector calculated by multiplying the window width (3, 5, ...) by the number of sensors. – [0023]), and
register the generated M-dimensional vector at each time point in a database as a normal vector (normal data is stored as training data – [0008]; learning data stored – [0013]; Next, the clustering unit 105 sets the initial positions of the cluster centers based on the extracted feature vectors (S304), performs clustering (S305), and adjusts the members of each cluster (S306). Next, the learning data storage unit 106 records the center and cluster members of each cluster (S307) – [0021]; Furthermore, past sensor signals 102 are assumed to be stored in a database, associated with the equipment ID and time. - [0052]); and
a normal vector distribution density leveling unit configured to divide a vector space of the normal vectors, select a predetermined number of normal vectors from the normal vectors included in each divided space so as to level a distribution density of the normal vectors, and
optimize a division number of the vector space of the normal vectors such that a total number of the selected normal vectors becomes a predesignated number (In step S304, the initial position of the cluster center is set. The processing flow will be explained using Figure 4.
First, specify the number of clusters (S401). Next, the first feature vector of the specified training period is set as the first cluster center (S402). Next, the similarity between the configured cluster centers and all feature vectors during the learning period is calculated (S403). Next, the maximum similarity score with the cluster center is calculated for all feature vectors (S404), and the feature vector with the smallest similarity score is designated as the next cluster center (S405). In other words, the cluster center is defined as the feature vector furthest from the nearest cluster center. In the first processing, the nearest cluster center will be the cluster center set in S402. Next, in step S406, if the number of clusters has reached the number of clusters specified in S401 (if YES in S406), the process is terminated (S407).
Otherwise (if NO in S406), repeat the process from S403 to S406 – [0025]; Figure 6 illustrates the flow of the process for dividing the cluster obtained by the above process so that the number of cluster members is less than or equal to a predetermined number - [0028]; Next, the processing flow in the cluster selection unit 107, the normal model creation unit 108, the abnormality measure calculation unit 109, and the threshold calculation unit 110 will be explained using Figure 8. In the learning data storage unit 106, the cluster center and members are recorded by the process shown in step S307 of Figure 3 (S801). A group number for cross-validation is assigned to all feature vectors for the specified training period (S802). Groups can be determined by specifying a period, such as designating one day as one group, or by dividing the data equally among a predetermined number of groups - [0032]).”
In regards to Claim 2, Shibuya discloses the claimed invention as detailed above. Shibuya further teaches “wherein the normal vector distribution density leveling unit is configured to equally divide the vector space into a plurality of cells, and select a predetermined number of normal vectors from each cell including normal vectors so as to level normal vector distribution density, and optimize sizes of the cells such that a total number of the selected normal vectors becomes a predesignated number (In step S304, the initial position of the cluster center is set. The processing flow will be explained using Figure 4. First, specify the number of clusters (S401). Next, the first feature vector of the specified training period is set as the first cluster center (S402). Next, the similarity between the configured cluster centers and all feature vectors during the learning period is calculated (S403). Next, the maximum similarity score with the cluster center is calculated for all feature vectors (S404), and the feature vector with the smallest similarity score is designated as the next cluster center (S405). In other words, the cluster center is defined as the feature vector furthest from the nearest cluster center. In the first processing, the nearest cluster center will be the cluster center set in S402. Next, in step S406, if the number of clusters has reached the number of clusters specified in S401 (if YES in S406), the process is terminated (S407). Otherwise (if NO in S406), repeat the process from S403 to S406 – [0025]; Figure 6 illustrates the flow of the process for dividing the cluster obtained by the above process so that the number of cluster members is less than or equal to a predetermined number - [0028]; Figure 6 illustrates the flow of the process for dividing the cluster obtained by the above process so that the number of cluster members is less than or equal to a predetermined number. However, this process does not need to be performed. First, specify the number of cluster members (S601). The system focuses on the first cluster (S602), compares the number of members in the cluster with the specified number of members (S603), and if the number of members in the cluster is less than the specified number of members (if the answer is YES in S603), the cluster is considered processed (S604). If all clusters have been processed (if the answer is YES in S605), the process will terminate (S609) – [0024]; Groups can be determined by specifying a period, such as designating one day as one group, or by dividing the data equally among a predetermined number of groups. - [0032]).”
In regards to Claim 3, Shibuya discloses the claimed invention as detailed above. Shibuya further teaches “wherein the normal vector distribution density leveling unit is configured to set, within the vector space, a plurality of spheres whose center positions do not overlap each other and whose center positions are defined as randomly selected normal vectors, select a predetermined number of normal vectors from each sphere so as to level the distribution density of the normal vectors, and optimize the sizes of the spheres such that the total number of the selected normal vectors becomes a predesignated number (In step S304, the initial position of the cluster center is set. The processing flow will be explained using Figure 4. First, specify the number of clusters (S401). Next, the first feature vector of the specified training period is set as the first cluster center (S402). Next, the similarity between the configured cluster centers and all feature vectors during the learning period is calculated (S403). Next, the maximum similarity score with the cluster center is calculated for all feature vectors (S404), and the feature vector with the smallest similarity score is designated as the next cluster center (S405). In other words, the cluster center is defined as the feature vector furthest from the nearest cluster center. In the first processing, the nearest cluster center will be the cluster center set in S402. Next, in step S406, if the number of clusters has reached the number of clusters specified in S401 (if YES in S406), the process is terminated (S407). Otherwise (if NO in S406), repeat the process from S403 to S406 – [0025]; Figure 6 illustrates the flow of the process for dividing the cluster obtained by the above process so that the number of cluster members is less than or equal to a predetermined number - [0028]; Next, the processing flow in the cluster selection unit 107, the normal model creation unit 108, the abnormality measure calculation unit 109, and the threshold calculation unit 110 will be explained using Figure 8. In the learning data storage unit 106, the cluster center and members are recorded by the process shown in step S307 of Figure 3 (S801). A group number for cross-validation is assigned to all feature vectors for the specified training period (S802). Groups can be determined by specifying a period, such as designating one day as one group, or by dividing the data equally among a predetermined number of groups - [0032]).”
In regards to Claim 6, Shibuya discloses the claimed invention as detailed above in Claim 1. Shibuya further teaches “clip a time-series signal in a predetermined period from M ( 2) or more types of time-series signals indicating an operation state of a facility during an abnormality monitoring period of the facility (Therefore, multiple sensors are installed on the target equipment or plant, and it is determined whether they are functioning normally or abnormally according to the monitoring criteria for each sensor. – [0004]; To solve the above problems, the present invention provides a method for monitoring the state of equipment or device by extracting feature vectors from sensor signals output from multiple sensors attached to the equipment or device, accumulating the center of each cluster and the feature vectors belonging to each cluster obtained by clustering these extracted feature vectors as training data, extracting feature vectors from new sensor signals output from multiple sensors attached to the equipment or device, selecting a cluster corresponding to the feature vector extracted from this new sensor signal from the clusters accumulated as training data, selecting a predetermined number of feature vectors from the feature vectors belonging to the selected cluster according to the feature vector extracted from the new sensor signal, creating a normal model using this selected predetermined number of feature vectors, calculating an abnormality measure based on the newly observed feature vectors and the created normal model, and determining whether the state of the equipment or device is abnormal or normal based on the calculated abnormality measure – [0010]; The equipment 101 subject to condition monitoring includes equipment and plants such as gas turbines and steam turbines. Equipment 101 outputs a sensor signal 102 that indicates its state. The sensor signal 102 is stored in the sensor signal storage unit 103. Figure 2 shows an example of listing the sensor signals 102 and representing them in a table format. The sensor signal 102 is a multidimensional time-series signal acquired at regular intervals. The table listing it consists of a date and time column 201 and data columns 202 for multiple sensor values provided on the equipment 101, as shown in Figure 2. The number of sensors can range from hundreds to thousands, and they may include sensors for the temperature of cylinders, oil, and coolant, the pressure of oil and coolant, the rotational speed of the shaft, room temperature, and operating time. In addition to representing outputs and states, control signals can also be used to control something to a specific value - [0019]; As a minimum measure, it is necessary to remove sensor signals with very low variance and sensor signals that are monotonically increasing – [0023]; Additionally, it's possible to remove invalid signals using correlation analysis. This method involves performing correlation analysis on multidimensional time-series signals. If there are multiple signals with a correlation value close to 1, indicating a high degree of similarity, these signals are considered redundant. The redundant signals are then removed, leaving only the non-redundant ones - [0024]);
generate, from the clipped time-series signal, an M-dimensional vector including M types of variables at a same time point as an abnormality determination target vector (First, the processing flow in the feature vector extraction unit 104, the clustering unit 105, and the training data storage unit 106 will be explained using Figure 3. First, the feature vector extraction unit 104 receives sensor signals 102 from the sensor signal storage unit 103 for a period specified as the learning period (S301), canonizes each sensor signal (S302), and then extracts feature vectors (S303). - [0021]; In step S303, feature vector extraction is performed for each time step. One approach is to simply list the canonicalized sensor signals, but it is also possible to create windows of ±1, ±2, ... for a given time point and extract features representing the time evolution of the data using a feature vector calculated by multiplying the window width (3, 5, ...) by the number of sensors. – [0023]); and
conduct an abnormality determination of the facility based on a distance between the generated abnormality determination target vector and a normal vector registered in a database by the normal vector registration device according to claim 1 (Furthermore, in order to solve the above problems, the present invention provides a method for monitoring the status of equipment, which includes the steps of creating and storing learning data based on sensor signals output from a plurality of sensors attached to the equipment or device, and identifying abnormalities in the sensor signals newly output from the plurality of sensors attached to the equipment or device, wherein the step of creating and storing learning data includes the steps of performing mode division according to operating state based on event signals output from the equipment or device, extracting feature vectors from the sensor signals, clustering the extracted feature vectors, storing the center of each cluster obtained by this clustering and the feature vectors belonging to the cluster as learning data, selecting one or several clusters from the clusters stored as learning data according to the extracted feature vector for each extracted feature vector, and selecting a predetermined number from the feature vectors belonging to the selected clusters according to the extracted feature vector, and selecting a predetermined number The method for identifying abnormalities in sensor signals includes the steps of creating a normal model using feature vectors belonging to selected clusters, calculating an anomaly measure based on the extracted feature vectors and the normal model, and calculating a threshold for each mode based on the calculated anomaly measure. The method for identifying abnormalities in sensor signals includes the steps of performing mode division according to operating state based on event signals, extracting feature vectors from newly observed sensor signals, selecting one or several clusters from clusters accumulated as training data according to newly observed feature vectors, selecting a predetermined number of feature vectors from those selected clusters according to the newly observed feature vectors, creating a normal model using the feature vectors belonging to the selected clusters, calculating an anomaly measure based on the newly observed feature vectors and the normal model, and determining whether the sensor signal is abnormal or normal based on the calculated anomaly measure, the mode, and the threshold calculated for each mode. - [0012]; In the normal model creation unit 108, a predetermined number of feature vectors are selected from among the feature vectors of groups that are members of the selected cluster but are different from the vector of interest, in order of proximity to the vector of interest (S805), and a normal model is created using these feature vectors (S806). The anomaly measure calculation unit 109 calculates an anomaly measure based on the distance of the feature vector of interest to the normal model (S807). Finally, it is checked whether the calculation of abnormality measures for all vectors has been completed (S808). If the calculation of abnormality measures for all vectors has been completed (if the answer is YES in S808), the threshold calculation unit 110 sets a threshold based on the abnormality measures of all vectors (S810). If it is determined in step S808 that not all anomaly measure calculations have been completed (the answer is NO in S808), then the next feature vector is considered (S809), and the process from steps S804 to S808 is repeated - [0033]).”
In regards to Claim 8, Shibuya discloses the claimed invention as detailed above in claim 1. Shibuya further teaches “clipping a time-series signal in a predetermined period from M ( 2) or more types of time-series signals indicating an operation state of a facility during an abnormality monitoring period of the facility (Therefore, multiple sensors are installed on the target equipment or plant, and it is determined whether they are functioning normally or abnormally according to the monitoring criteria for each sensor. – [0004]; To solve the above problems, the present invention provides a method for monitoring the state of equipment or device by extracting feature vectors from sensor signals output from multiple sensors attached to the equipment or device, accumulating the center of each cluster and the feature vectors belonging to each cluster obtained by clustering these extracted feature vectors as training data, extracting feature vectors from new sensor signals output from multiple sensors attached to the equipment or device, selecting a cluster corresponding to the feature vector extracted from this new sensor signal from the clusters accumulated as training data, selecting a predetermined number of feature vectors from the feature vectors belonging to the selected cluster according to the feature vector extracted from the new sensor signal, creating a normal model using this selected predetermined number of feature vectors, calculating an abnormality measure based on the newly observed feature vectors and the created normal model, and determining whether the state of the equipment or device is abnormal or normal based on the calculated abnormality measure – [0010]; The equipment 101 subject to condition monitoring includes equipment and plants such as gas turbines and steam turbines. Equipment 101 outputs a sensor signal 102 that indicates its state. The sensor signal 102 is stored in the sensor signal storage unit 103. Figure 2 shows an example of listing the sensor signals 102 and representing them in a table format. The sensor signal 102 is a multidimensional time-series signal acquired at regular intervals. The table listing it consists of a date and time column 201 and data columns 202 for multiple sensor values provided on the equipment 101, as shown in Figure 2. The number of sensors can range from hundreds to thousands, and they may include sensors for the temperature of cylinders, oil, and coolant, the pressure of oil and coolant, the rotational speed of the shaft, room temperature, and operating time. In addition to representing outputs and states, control signals can also be used to control something to a specific value - [0019]; As a minimum measure, it is necessary to remove sensor signals with very low variance and sensor signals that are monotonically increasing – [0023]; Additionally, it's possible to remove invalid signals using correlation analysis. This method involves performing correlation analysis on multidimensional time-series signals. If there are multiple signals with a correlation value close to 1, indicating a high degree of similarity, these signals are considered redundant. The redundant signals are then removed, leaving only the non-redundant ones - [0024]);
generating, from the clipped time-series signal, an M-dimensional vector including M types of variables at a same time point as an abnormality determination target vector (First, the processing flow in the feature vector extraction unit 104, the clustering unit 105, and the training data storage unit 106 will be explained using Figure 3. First, the feature vector extraction unit 104 receives sensor signals 102 from the sensor signal storage unit 103 for a period specified as the learning period (S301), canonizes each sensor signal (S302), and then extracts feature vectors (S303). - [0021]; In step S303, feature vector extraction is performed for each time step. One approach is to simply list the canonicalized sensor signals, but it is also possible to create windows of ±1, ±2, ... for a given time point and extract features representing the time evolution of the data using a feature vector calculated by multiplying the window width (3, 5, ...) by the number of sensors. – [0023]); and
conducting an abnormality determination of the facility based on a distance between the generated abnormality determination target vector and a normal vector registered in a database by the normal vector registration device according to claim 1 (Furthermore, in order to solve the above problems, the present invention provides a method for monitoring the status of equipment, which includes the steps of creating and storing learning data based on sensor signals output from a plurality of sensors attached to the equipment or device, and identifying abnormalities in the sensor signals newly output from the plurality of sensors attached to the equipment or device, wherein the step of creating and storing learning data includes the steps of performing mode division according to operating state based on event signals output from the equipment or device, extracting feature vectors from the sensor signals, clustering the extracted feature vectors, storing the center of each cluster obtained by this clustering and the feature vectors belonging to the cluster as learning data, selecting one or several clusters from the clusters stored as learning data according to the extracted feature vector for each extracted feature vector, and selecting a predetermined number from the feature vectors belonging to the selected clusters according to the extracted feature vector, and selecting a predetermined number The method for identifying abnormalities in sensor signals includes the steps of creating a normal model using feature vectors belonging to selected clusters, calculating an anomaly measure based on the extracted feature vectors and the normal model, and calculating a threshold for each mode based on the calculated anomaly measure. The method for identifying abnormalities in sensor signals includes the steps of performing mode division according to operating state based on event signals, extracting feature vectors from newly observed sensor signals, selecting one or several clusters from clusters accumulated as training data according to newly observed feature vectors, selecting a predetermined number of feature vectors from those selected clusters according to the newly observed feature vectors, creating a normal model using the feature vectors belonging to the selected clusters, calculating an anomaly measure based on the newly observed feature vectors and the normal model, and determining whether the sensor signal is abnormal or normal based on the calculated anomaly measure, the mode, and the threshold calculated for each mode. - [0012]; In the normal model creation unit 108, a predetermined number of feature vectors are selected from among the feature vectors of groups that are members of the selected cluster but are different from the vector of interest, in order of proximity to the vector of interest (S805), and a normal model is created using these feature vectors (S806). The anomaly measure calculation unit 109 calculates an anomaly measure based on the distance of the feature vector of interest to the normal model (S807). Finally, it is checked whether the calculation of abnormality measures for all vectors has been completed (S808). If the calculation of abnormality measures for all vectors has been completed (if the answer is YES in S808), the threshold calculation unit 110 sets a threshold based on the abnormality measures of all vectors (S810). If it is determined in step S808 that not all anomaly measure calculations have been completed (the answer is NO in S808), then the next feature vector is considered (S809), and the process from steps S804 to S808 is repeated - [0033])”
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 4-5 are rejected under 35 U.S.C. 103 as being unpatentable over Shibuya in view of Hara (JP2020027342A).
In regards to Claim 4, Shibuya discloses the claimed invention as detailed above in Claim 1. Shibuya further teaches “a normal vector selection unit configured to: calculate a distance between each normal vector registered in the database and another normal vector registered in the database (Next, the distance between all feature vectors for the specified learning period and each cluster center vector is calculated - [0027]); and delete the normal vector from the database (Additionally, it's possible to remove invalid signals using correlation analysis. This method involves performing correlation analysis on multidimensional time-series signals. If there are multiple signals with a correlation value close to 1, indicating a high degree of similarity, these signals are considered redundant. The duplicate signals are then removed, leaving only the non-duplicate ones - [0024]).”
Shibuya is silent with regards to the language of “a normal vector selection unit configured to: calculate a distance between each normal vector registered in the database and another normal vector registered in the database; and delete the normal vector from the database in accordance with the calculated distance”
Hara teaches “a normal vector selection unit configured to: calculate a distance between each normal vector registered in the database and another normal vector registered in the database (In step S2001, the determination unit 12 calculates the distance between each of the training data. Here, the Euclidean distance between one normal data point, which is an n-dimensional vector, and each of the other normal data points is calculated for each normal data point used as training data. - [0066]); and delete the normal vector from the database in accordance with the calculated distance (Next, the determination unit 12 removes a predetermined number of normal data from the training data based on each distance (step S1002). This removes similar, normal data. Here, for example, the process of removing one of two normal data points in order of the shortest distance between them may be repeated. This is because two normal data points with a short distance between them are considered to be similar - [0067])”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Shibuya to incorporate the teaching of Harat o utilize the distance between the normal vectors to delete normal vectors that are similar to each other. By deleting similar normal vectors this is an improvement that yields predictable results in the speed and efficiency of the evaluation of normal and abnormal data by reducing the amount of data that has to be evaluated.
In regards to Claim 5, Shibuya in view of Hara discloses the claimed invention as detailed above. Shibuya further teaches “wherein the normal vector selection unit is configured to perform processing including: calculating a distance between each normal vector registered in the database and another normal vector registered in the database (Next, the distance between all feature vectors for the specified learning period and each cluster center vector is calculated - [0027]).”
Shibuya is silent with regards to the language of “extracting a predetermined number of the other normal vectors as neighboring vectors in ascending order of the calculated distance; and deleting the normal vector from the database in accordance with a distance to a centroid vector of the extracted neighboring vector”
Hara further teaches “extracting a predetermined number of the other normal vectors as neighboring vectors in ascending order of the calculated distance; and deleting the normal vector from the database in accordance with a distance to a centroid vector of the extracted neighboring vector (Alternatively, the determination unit 12 may repeatedly remove normal data starting from the region with the largest number of normal data, in order [i.e. ascending order] of the shortest distance between two normal data points in that region. The example in Figure 10 shows an example of the number of normal data points in each of the nine regions 1001 to 1009 in two dimensions, where each normalized normal data point is contained. In the example shown in Figure 10, the determination unit 12 may remove one of the two normal data points, 1010 and 1011, from the region 1004 that contains the most normal data points, where the distance d between the two normal data points is shortest. The determination unit 12 may then repeat the same process until a predetermined number of normal data points are removed. - [0073])”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Shibuya to incorporate the teaching of Harat o utilize the distance between the normal vectors to delete normal vectors that are similar to each other. By deleting similar normal vectors this is an improvement that yields predictable results in the speed and efficiency of the evaluation of normal and abnormal data by reducing the amount of data that has to be evaluated.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Shibuya in view of Aizawa (JP2010015206A).
In regards to Claim 7, Shibuya discloses the claimed invention as detailed above in Claim 6. Shibuya is silent with regards to the language of “wherein the abnormality determination unit is configured to determine whether to repair the facility based on the number of times the facility has been determined to be abnormal.”
Aizawa teaches “wherein the abnormality determination unit is configured to determine whether to repair the facility based on the number of times the facility has been determined to be abnormal (This database stores the area where each piece of equipment is installed, the installation location of the equipment, the importance of the equipment, equipment measurement information, the level of the measurement information (measurement value level), the number of times the measurement information was judged to be abnormal (number of abnormal level measurements), the duration for which the abnormal level in the measurement information persisted (duration of abnormal level measurement signal), the equipment operating status, the equipment configuration (redundant configuration/single configuration), the corresponding equipment in the case of a redundant configuration (this item is set when the equipment importance is A), connected equipment (information about the equipment connected to each piece of equipment is stored (this information changes depending on the system status)), and priority (t) for each area equipment importance – [0022]; Next, in Step 4-3, for each equipment status acquisition cycle, if the level of the measured value obtained in Step 4-2 is detected to be within the range of abnormal measured value levels defined in DB123, it is counted as an abnormal level being detected and registered in the number of abnormal level measurements in DB121 (if the previous measurement was within the normal range).If an abnormal level was measured previously, the measurement time for the abnormal level of the DB121 signal will be updated because the abnormal level is continuing – [0037]; In Step 4-4, for equipment where an anomaly was detected in Step 4-3, DB121 is referenced. For equipment with a redundant configuration (severity A), the measurement level of the corresponding equipment constituting the redundant system is also checked, and the process proceeds to Step 4-5 to update the priority – [0038]; If an abnormality is detected in any of these pieces of equipment, it is necessary to prioritize the repair and inspection of the equipment in question. In this case, the priority order for equipment repair and inspection will be determined in the following two stages – [0120]).”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify Shibuya to incorporate the teaching of Aizawa to utilize a priority system with the number of abnormalities measured to determine when maintenance should occur. By utilizing a priority system with maintenance this is an improvement to the addressing failures such that it minimizes the damage to society and to keep the systems in operation.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to YOSSEF KORANG-BEHESHTI whose telephone number is (571)272-3291. The examiner can normally be reached Monday - Friday 10:00 am - 6:30 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Catherine Rastovski can be reached at (571) 270-0349. 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.
/YOSSEF KORANG-BEHESHTI/Primary Examiner, Art Unit 2857
Prosecution Insights