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 .
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 16-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because the recited “computer program product comprising instructions” amounts to the recitation of “a computer program per se (often referred to as "software per se") [that is] claimed as a product without any structural recitations” [See MPEP 2106.03 – “Non-limiting examples of claims that are not directed to any of the statutory categories include: Products that do not have a physical or tangible form, such as … a computer program per se (often referred to as "software per se") when claimed as a product without any structural recitations”].
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 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)(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.
Claim(s) 1, 3-5, 9-13, 16, and 18-20 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Wen et al. (US 20220201026 A1)[hereinafter “Wen”].
Regarding Claims 1 and 16, Wen discloses a method (and corresponding computer program product containing computer instructions [Paragraph [0005]]) for training a diagnostic model for diagnosing a production system, wherein the production system comprises a plurality of sub-systems [Abstract – “Examples of techniques for threat detection in an industrial process system are described herein. An aspect includes determining a plurality of subsystems of an industrial process system. Another aspect includes, for each of the plurality of subsystems, constructing and training a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system. Another aspect includes monitoring the industrial process system using the plurality of DAE models corresponding to the plurality of subsystems. Another aspect includes, based on the plurality of DAE models, determining a cyberattack in a subsystem of the plurality of subsystems.”Paragraph [0012] – “Examples of an industrial process system that may be connected to a network include but are not limited to a manufacturing system or a power generation system.”],
the diagnostic model comprising:
for each sub-system [See Fig. 6 and Paragraph [0039] – “Each subsystem input data of subsystem input data 601A-N is fed into a respective DAE model of DAE models 602A-N corresponding to the same subsystem.”], a corresponding first learning model arranged to receive input data, and to generate compressed data for the corresponding sub-system in a corresponding latent space [See Fig. 5 (the levels after compression being the latent space) and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes first level compression 502 … [and] second level reconstruction 506”]; and
a second learning model arranged to receive the compressed data generated by the first learning models, and generate further compressed data for the production system in a further latent space [See Fig. 5 (the levels after compression being the latent space) and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes … second level compression 503 … [and] first level reconstruction 505”];
the method comprising training, by a hardware computer system, of the first and second learning models based on training data derived from sensor data characterizing the sub-systems [Paragraph [0003] – “The processor can also, for each of the plurality of subsystems, construct and train a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system.” See also Paragraph [0005].Paragraph [0028] – “The training data that is used in block 203 to train the plurality of DAE models may, in various embodiments, be collected from the industrial process itself”See the sensors of Fig. 3 and Paragraph [0031] and Fig. 4, Paragraph [0033] – “FIG. 4 is a block diagram of an example node graph 400 of an industrial process system, such as system 300 of FIG. 3, for use in conjunction with threat detection in an industrial process system. … Each node of nodes 41A-G, 42A-C, and 43A-C may correspond to a respective IT or OT element of the industrial process system (for example, a computer processing device, a valve, a sensor, or an actuator).”].
Regarding Claims 3 and 18, Wen discloses a diagnostic method for diagnosing a subject production system which comprises a plurality of sub-systems, using the diagnostic model trained by the method according to claim 1 [Abstract – “Examples of techniques for threat detection in an industrial process system are described herein. An aspect includes determining a plurality of subsystems of an industrial process system. Another aspect includes, for each of the plurality of subsystems, constructing and training a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system. Another aspect includes monitoring the industrial process system using the plurality of DAE models corresponding to the plurality of subsystems. Another aspect includes, based on the plurality of DAE models, determining a cyberattack in a subsystem of the plurality of subsystems.”], the diagnostic method comprising:
for each sub-system [See Fig. 6 and Paragraph [0039] – “Each subsystem input data of subsystem input data 601A-N is fed into a respective DAE model of DAE models 602A-N corresponding to the same subsystem.”], inputting data derived from sensor data characterizing the sub-system of the subject production system [See the sensors of Fig. 3 and Paragraph [0031] and Fig. 4, Paragraph [0033] – “FIG. 4 is a block diagram of an example node graph 400 of an industrial process system, such as system 300 of FIG. 3, for use in conjunction with threat detection in an industrial process system. … Each node of nodes 41A-G, 42A-C, and 43A-C may correspond to a respective IT or OT element of the industrial process system (for example, a computer processing device, a valve, a sensor, or an actuator).”Paragraph [0036] – “Elements 51A-N in input data 501 may each correspond to a respective node in a subsystem of an industrial process system, such as nodes 41A-G of cluster 401 of FIG. 4.”] into the corresponding first learning model to generate corresponding compressed data for the sub-system of the subject production system [See Fig. 5 and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes first level compression 502 … [and] second level reconstruction 506”];
inputting the compressed data generated by the first learning models into the second learning model to generate corresponding further compressed data for the subject production system [See Fig. 5 and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes … second level compression 503 … [and] first level reconstruction 505”]; and
diagnosing the subject production system based on data generated by at least one selected from: (i) at least one first learning model of the first learning models and/or (ii) the second learning model [Paragraph [0036] – “A difference, or error, between the input data 501 and the reconstructed data 507 may be used to detect anomalies in a subsystem of an industrial process system that corresponds to the DAE model 500.”].
Regarding Claims 4 and 19, Wen discloses that at least one learning model of the first and second learning models is a representation learning model [See Fig. 5 and Paragraph [0010] – “FIG. 5 is a block diagram of an example deep autoencoder (DAE) model for use in conjunction with threat detection in an industrial process system”Paragraph [0018] of the instant Specification discloses that an autoencoder is a type of representation learning model.] or a manifold learning model.
Regarding Claim 5, Wen discloses that at least one learning model of the first and second learning models is an autoencoder [See Fig. 5 and Paragraph [0010] – “FIG. 5 is a block diagram of an example deep autoencoder (DAE) model for use in conjunction with threat detection in an industrial process system”Paragraph [0018] of the instant Specification discloses that an autoencoder is a type of representation learning model.].
Regarding Claim 9, Wen discloses that the production system is the subject production system at an earlier time [Fig. 2, training step 203 happens prior to use of the models to determine a possible cyber attack].
Regarding Claim 10, Wen discloses that diagnosing the subject production system comprises determining whether an anomaly score of the data generated by the second learning model is above a predetermined threshold [Paragraph [0029] – “In block 205, an error in the reconstructed data that is output by each of the DAE models is determined. In block 206, it is determined if any of the DAE models has an error that is higher than a threshold. … In block 207, based on the error being above the threshold, it is determined that there is an anomaly, which may correspond to a cyberattack, in the one or more subsystems corresponding to the one or more DAE models that were determined to have an error higher than the threshold in block 206.”].
Regarding Claim 11, Wen discloses that the second learning model is an autoencoder [See Fig. 5 and Paragraph [0010] – “FIG. 5 is a block diagram of an example deep autoencoder (DAE) model for use in conjunction with threat detection in an industrial process system”], and the anomaly score is indicative of a discrepancy between the compressed data generated by the first learning models and reconstructed data generated by the second learning model upon receiving the compressed data generated by the first learning models [Paragraph [0036] – “A difference, or error, between the input data 501 and the reconstructed data 507 may be used to detect anomalies in a subsystem of an industrial process system that corresponds to the DAE model 500.”].
Regarding Claim 12, Wen discloses that at least one first learning model of the first learning models is an autoencoder [See Fig. 5 and Paragraph [0010] – “FIG. 5 is a block diagram of an example deep autoencoder (DAE) model for use in conjunction with threat detection in an industrial process system”], and the anomaly score is indicative of a discrepancy between the data input to the at least one first learning model of the first learning models which is an autoencoder, and reconstructed data generated by the at least one first learning model of the first learning models [Paragraph [0036] – “A difference, or error, between the input data 501 and the reconstructed data 507 may be used to detect anomalies in a subsystem of an industrial process system that corresponds to the DAE model 500.”].
Regarding Claim 13, Wen discloses that the anomaly score is based on data in at least one of the latent spaces [See Fig. 5 (the levels after compression being the latent space) and Paragraph [0036] – “A difference, or error, between the input data 501 and the reconstructed data 507 may be used to detect anomalies in a subsystem of an industrial process system that corresponds to the DAE model 500.”].
Regarding Claim 20, Wen discloses a non-transitory computer program product comprising instructions therein, which instructions, when executed by a computer system, are configured to cause the computer system to [Paragraph [0005]] at least:
for each sub-system of a plurality of sub-systems of a subject production system [See Fig. 6 and Paragraph [0039] – “Each subsystem input data of subsystem input data 601A-N is fed into a respective DAE model of DAE models 602A-N corresponding to the same subsystem.”], input data derived from sensor data characterizing the sub-system of the subject production system [See the sensors of Fig. 3 and Paragraph [0031] and Fig. 4, Paragraph [0033] – “FIG. 4 is a block diagram of an example node graph 400 of an industrial process system, such as system 300 of FIG. 3, for use in conjunction with threat detection in an industrial process system. … Each node of nodes 41A-G, 42A-C, and 43A-C may correspond to a respective IT or OT element of the industrial process system (for example, a computer processing device, a valve, a sensor, or an actuator).”Paragraph [0036] – “Elements 51A-N in input data 501 may each correspond to a respective node in a subsystem of an industrial process system, such as nodes 41A-G of cluster 401 of FIG. 4.”] into a corresponding first trained learning model to generate corresponding compressed data for the sub-system of the subject production system in a corresponding latent space [See Fig. 5 (the levels after compression being the latent space) and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes first level compression 502 … [and] second level reconstruction 506”];
input the compressed data generated by the first trained learning models into a second trained learning model to generate corresponding further compressed data for the subject production system in a further latent space [See Fig. 5 (the levels after compression being the latent space) and Paragraph [0036] – “FIG. 5 is a block diagram of a DAE model 500 for threat detection in an industrial process system. … Model 500 includes … second level compression 503 … [and] first level reconstruction 505”]; and
diagnose the subject production system based on data generated by at least one selected from: (i) at least one first trained learning model of the first trained learning models and/or (ii) the second trained learning model [Paragraph [0036] – “A difference, or error, between the input data 501 and the reconstructed data 507 may be used to detect anomalies in a subsystem of an industrial process system that corresponds to the DAE model 500.”].
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.
Claim(s) 2 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wen et al. (US 20220201026 A1)[hereinafter “Wen”].
Regarding Claims 2 and 17, Wen discloses that at least one sub-system of the plurality of sub-systems comprises a plurality of sub-sub-systems [See Fig. 4 and Paragraph [0033] – “Cluster 401 comprises a first subsystem of the industrial process system, and includes nodes 41A-G. Nodes 41A-B in cluster 401 are directly connected to nodes in clusters 402 and 403, while the remaining nodes 41C-G are only directly connected to nodes within cluster 401. Cluster 402, corresponding to a second subsystem of the industrial process system, includes nodes 42A-C. Cluster 403, corresponding to a third subsystem of the industrial process system, includes nodes 43A-C. Each node of nodes 41A-G, 42A-C, and 43A-C may correspond to a respective IT or OT element of the industrial process system (for example, a computer processing device, a valve, a sensor, or an actuator).” The clusters being subsystems and their nodes being subsubsystems.].
Wen fails to explicitly disclose the third learning model to generate compressed data or reconstruction. However, Wen does disclose that different numbers of levels of compression and levels of reconstruction were contemplated [Paragraph [0038] – “FIG. 5 is shown for illustrative purposes only. For example, a DAE model such as is shown in FIG. 5 may have any appropriate number of levels of compression and levels of reconstruction, each level having any appropriate number of elements.”]. It would have been obvious to extend the DAE to include an additional and initial pair of compression/reconstruction layers for the subsubsystems because doing so would have amounted to a contemplated duplication of parts (see MPEP 2144.04(VI)(B) – “mere duplication of parts has no patentable significance unless a new and unexpected result is produced.”).
Wen, as modified, would disclose that the diagnostic model further comprises, for each sub-sub-system [Paragraph [0036] – “Elements 51A-N in input data 501 may each correspond to a respective node in a subsystem of an industrial process system, such as nodes 41A-G of cluster 401 of FIG. 4.”], a corresponding third learning model arranged to receive input data which is sensor data for the corresponding sub-sub-system, and to generate sub-sub-system compressed data in a corresponding latent space [The additional, initial third compression level and corresponding reconstruction level]; and
the input data received by the corresponding first learning model is the sub-sub- system compressed data for each sub-sub-system [Passing of the initial third compression level data to compression level 502].
Claim(s) 6-8 and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wen et al. (US 20220201026 A1)[hereinafter “Wen”] and Yu et al. (US 20220284166 A1)[hereinafter “Yu”].
Regarding Claim 6, Wen fails to disclose that at least one learning model of the first and second learning models apply a principal component analysis to input data.
However, Yu discloses using such a technique when performing machine learning in the context of a lithography context [See Paragraphs [0032]-[0035]. Paragraph [0035] – “principal component analysis (PCA)”]. It would have been obvious to employ such a technique because doing so would have been effective in performing machine learning.
Regarding Claim 7, Wen fails to disclose that at least one learning model of the first and second learning models is a t-distributed stochastic neighbour embedding.
However, Yu discloses using such a technique when performing machine learning in the context of a lithography context [See Paragraphs [0032]-[0035]. Paragraph [0035] – “t-distributed stochastic neighbor embedding (t-sne)”]. It would have been obvious to employ such a technique because doing so would have been effective in performing machine learning.
Regarding Claim 8, Wen fails to disclose that at least one learning model of the first and second learning models is a Uniform Manifold Approximation.
However, Yu discloses using such a technique when performing machine learning in the context of a lithography context [See Paragraphs [0032]-[0035]. Paragraph [0035] – “uniform manifold approximation and projection (u-map)”]. It would have been obvious to employ such a technique because doing so would have been effective in performing machine learning.
Regarding Claim 15, Wen fails to disclose that the production system is a lithographic apparatus.
However, Yu discloses performing machine learning in the context of a lithography context [See Paragraphs [0032]-[0035]]. It would have been obvious to use machine learning to diagnose a lithography system as the production system in order to ensure its continued proper operation.
Claim(s) 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Wen et al. (US 20220201026 A1)[hereinafter “Wen”] and Hines et al. (US 20220383038 A1)[hereinafter “Hines”].
Regarding Claim 14, Wen fails to disclose that the anomaly score is indicative of a discrepancy between the compressed data and/or the further compressed data for the subject production system, and compressed data and/or further compressed data for a population of reference production systems.
However, Hines discloses the use of reference data in training a machine learning model in identifying an anomaly score in evaluating reconstruction error produced by an autoencoder [See Fig. 3 and Paragraph [0038]]. It would have been obvious to evaluate discrepancies between the compressed data for a subject production system relative to a population of reference production systems because doing so would have been an effective manner in determining if an anomaly is present.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
US 20220198279 A1 – Data-Driven Methodology For Automatic Detection Of Data Drift
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KYLE ROBERT QUIGLEY whose telephone number is (313)446-4879. The examiner can normally be reached 9AM-5PM EST.
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, Arleen Vazquez can be reached at (571) 272-2619. 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.
/KYLE R QUIGLEY/Primary Examiner, Art Unit 2857
Prosecution Insights