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 1-13 and 15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to abstract idea of a mental concept without significantly more. The claims recite the mental concept of providing one or more predictions to a test automation application. This judicial exception is not integrated into a practical application because the additional steps of receiving parameters, trained models, prediction requests, training data and a configuration file are mere data gather and storage. MPEP 2106.05(g). The additional step of “allow[ing] the machine learning system to access the training data to train at least one… neural networks…” doesn’t actually claim to train the neural network – so it amounts to mere data gathering as well – but if it did claiming training that would be insignificant extra solution activity. MPEP 2106.05(g) The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements directed to hardware are directed to generic computer parts. MPEP 2106.05(f).
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.
Claims 1-14 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 recites “configuration file t”, this seems to be a typo because it is not used again in the claims or explained in the specification.
Claim 4 recites “the one or more predictions metadata to the communication store.” There is insufficient antecedent basis for the predictions metadata.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-5, 10-11 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al and US20200226419A1 to Knaan et al.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al, US20200226419A1 to Knaan et al, and US20160245864A1 to Mydill.
Claims 7-9 are rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al, US20200226419A1 to Knaan et al, and US6832122B1 to Huber.
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al, US20200226419A1 to Knaan et al, and US 20150073751 A1 to Liao et al.
Claims 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir and US20220343141A1 to Lan et al.
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al, US6832122B1 to Huber and US20200226419A1 to Knaan et al.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over US20190370158A1 to Rivoir, US20220343141A1 to Lan et al and US6832122B1 to Huber.
Rivoir teaches claim 1. A manufacturing system, comprising:
a machine learning system, (Rivoir fig. 3 analysis 12) the machine learning system comprising:
one or more neural networks; and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…”)
a configuration file comprising information associated with a trained neural network for operations; (Rivoir para 318 “The model is returned as prediction variable and/or in a computer readable form as equations, table…” The file is the equations or table.)
a structured data store (Rivoir fig. 3 database 10) connected to the machine learning system, the structured data store having an interface to the machine learning system and an interface to a test automation application (Rivoir Fig. 3 test case generator 6 and test method.) used to test devices under test (DUTs); (Rivoir fig. 3 DUT 4) and
one or more processors configured to execute code to cause the one or more processors to control the structured data store to: (Rivoir para 352 “ Some or all of the method steps may be executed by … a microprocessor…”)
receive and store (Fig. 3 below)
PNG
media_image1.png
334
794
media_image1.png
Greyscale
receive and store reference parameters for the DUTs from the test automation application in a reference parameter store and to allow the machine learning system to access the reference parameters; (Rivoir para 27 “the test case generator is configured to randomly generate a plurality of test cases, wherein the test case generator is configured to randomly generate values for the one or more input variables (operating parameters) within an operating range of the device under test for the one or more input variables.”)
receive (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…” The prediction request is whatever causes the prediction of the variable combinations that will lead to an error.)
allow the machine learning system to store trained ones of the one or more neural networks in a trained models store in the structured data store; and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
allow the machine learning system to recall a selected trained neural network and an associated configuration file t and to provide one or more predictions to the test automation application. (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
Rivoir doesn’t teach in a training store within the structured data store and allow the machine learning system to access the training data to train at least one of the one or more neural networks in the machine learning system… receive and
However, Lan teaches in a training store within the structured data store and allow the machine learning system to access the training data to train at least one of the one or more neural networks in the machine learning system... (Lan abs “gathering state-action pair data from an expert policy; applying imitation learning to yield a cloned policy based on the gathered state-action pair data from the expert policy; and applying a reinforcement learning technique, wherein the reinforcement learning technique is initialized based on the cloned policy…” Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 11 “the cloned policy is in the form of a neural network, wherein the deepest hidden layer is convolutional in one dimension.”)
Rivoir, Lan and the claims are using machine learning models on diagnostic data. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to train using Lan’s training method because “it can outperform a well-trained expert model with a higher tuning success rate and fewer adjustment steps which leads to shorter total tuning time.” Lan para 9.
Rivoir doesn’t teach receive and store prediction requests…
However, Knaan teaches receive and store prediction requests… (Knaan fig. 3 and para 54 “trained machine learning model(s) 332 can receive input data 330 and one or more inference/prediction requests 340 (perhaps as part of input data 330) and responsively provide as an output one or more inferences and/or predictions 350.” The input data is in storage 330, so the inference requests are stored in input data storage 330.)
Rivoir, Knaan and the claims are using trained neural networks to make predictions. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to store prediction requests because the models are usually bottlenecks and requests shouldn’t be thrown away because they can’t immediately be run.
Lan teaches claim 2. The manufacturing system as claimed in claim 1, wherein the code that causes the one or more processors to control the structured data store to receive and store training data comprises code to cause the one or more processors to store input data and input metadata for training of the one or more of the neural networks. (Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 9 “the imitation and reinforcement learning based cavity filter tuning model of embodiments has been applied in a simulation environment and could tune cavity filters with more screws and satisfy both S11 and S21 parameters (return loss and insertion loss) and tuned both coupling and cross-coupling…” The signal is metadata because it is measurements on top of time.)
Rivoir teaches claim 3. The manufacturing system as claimed in claim 1, wherein the code that causes the one or more processors to control the structured data store to receive and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…” The prediction request is whatever causes the prediction of the variable combinations that will lead to an error.)
Rivoir doesn’t teach receive and store prediction requests…
However, Knaan teaches receive and store prediction requests… (Knaan fig. 3 and para 54 “trained machine learning model(s) 332 can receive input data 330 and one or more inference/prediction requests 340 (perhaps as part of input data 330) and responsively provide as an output one or more inferences and/or predictions 350.” The input data is in storage 330, so the inference requests are stored in input data storage 330.)
Rivoir teaches claim 4. The manufacturing system as claimed in claim 3, wherein the code that causes the one or more processor to control the structured data store to receive and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…” The prediction request is whatever causes the prediction of the variable combinations that will lead to an error.)
Rivoir doesn’t teach receive and store prediction requests…
However, Knaan teaches receive and store prediction requests… (Knaan fig. 3 and para 54 “trained machine learning model(s) 332 can receive input data 330 and one or more inference/prediction requests 340 (perhaps as part of input data 330) and responsively provide as an output one or more inferences and/or predictions 350.” The input data is in storage 330, so the inference requests are stored in input data storage 330.)
Rivoir teaches claim 5. The manufacturing system as claimed in claim 1, wherein the one or more processors are further configured to provide communication between the machine learning system and the test automation application. (Rivoir fig. 3 shows processors handing communication between test automation system (generator 6), through the test method and database to the ML system (analysis 12).)
Rivoir teaches claim 6. The manufacturing system as claimed in claim 1, wherein the structured data store is (Rivoir fig. 3 database.)
Rivoir doesn’t teach the replication.
However, Mydill teaches structured data store is replicated on a plurality of computing devices in the manufacturing system. (Mydill abs “Stimulus test signals of the data patterns are replicated and distributed to the devices.”)
Mydill, Rivoir and the claims are all directed to testing devices in a manufacturing setting. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to replicate the data across several devices because a “A common method of reducing the test cost of logic circuits in the DUT is to test multiple DUTs simultaneously…” Mydill para 5.
Rivoir teaches claim 7. The manufacturing system as claimed in claim 1, wherein the structured data store has (Rivoir fig. 3 database 10.)
Rivoir doesn’t have a global database.
However, Huber teaches the structured data store has a global portion stored on a global server accessible by a plurality of computing devices in the manufacturing system, and a local portion stored on the plurality of computing devices in the manufacturing system. (Huber fig. 1 global 20 and local 40. Huber claim 5 “global data used for a plurality of assembly lines and local data used at one of the plurality of assembly lines…”)
Rivoir, Huber and the claims all process data for manufacturing. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to have a global server that way “the differences between the data sets and the standardization data can be easily viewed using the difference editor.” Huber 2:15.
Lan teaches claim 8. The manufacturing system as claimed in claim 7, wherein the (Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 9 “the imitation and reinforcement learning based cavity filter tuning model of embodiments has been applied in a simulation environment and could tune cavity filters with more screws and satisfy both S11 and S21 parameters (return loss and insertion loss) and tuned both coupling and cross-coupling…”)
Lan doesn’t teach a global portion.
However, Huber teaches a global portion. (Huber fig. 1 global 20 and local 40. Huber claim 5 “global data used for a plurality of assembly lines and local data used at one of the plurality of assembly lines…”)
Rivoir, Lan, Huber and the claims all process data for manufacturing. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to have a global server that way “the differences between the data sets and the standardization data can be easily viewed using the difference editor.” Huber 2:15.
Rivoir teaches claim 9. The manufacturing system as claimed in claim 7, wherein the local portion comprises the communication store. (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…” Rivoir fig. 3 database 10.)
Rivoir teaches claim 10. The manufacturing system as claimed in claim 1, wherein at least the trained neural networks trained by data from the test automation application reside on a plurality of computing devices in the manufacturing system. (Rivoir fig. 1)
Lan teaches claim 11. The manufacturing system as claimed in claim 1, wherein at least one of the one or more processors reside in the machine learning system to execute code to train one or more of the neural networks and to deploy the one or more neural networks to make predictions for the test automation application. (Lan abs “gathering state-action pair data from an expert policy; applying imitation learning to yield a cloned policy based on the gathered state-action pair data from the expert policy; and applying a reinforcement learning technique, wherein the reinforcement learning technique is initialized based on the cloned policy…” Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 11 “the cloned policy is in the form of a neural network, wherein the deepest hidden layer is convolutional in one dimension.”)
Rivoir teaches claim 12. The manufacturing system as claimed in claim 11, wherein the at least one of the one or more processors is further configured to generate a tensor array and perform feature extraction on (Rivoir para 115 “Feature selection algorithms identify the smallest set of most influencing variables as candidates for which variables (y) to correct as a function of which other variables (x).” The tensor array is the list of variable x.)
Rivoir doesn’t teach training.
However, Liao teaches how to generate a tensor array and perform feature extraction on both the training data and the input data for predictions during deployment, providing consistency for both training and run time. (Liao para 53 “the AP applies the training algorithm (operation 478) to train a baseline base on the saved data array which contains all the extracted features. After a baseline is trained (decision 472), the AP applies the testing algorithm and outputs prediction results (operation 476) when new data arrives.”)
Liao, Rivoir and the claims all teach predicting errors in a manufacturing environment. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to train on the same type of data you eventually run on because that’s the default setup for machine learning.
Rivoir teaches claim 13. The manufacturing system as claimed in claim 1, wherein the structured data store comprises:
a reference parameter store to store reference parameters and waveforms associated with reference parameters acquired from the device under test; (Rivoir para 27 “the test case generator is configured to randomly generate a plurality of test cases, wherein the test case generator is configured to randomly generate values for the one or more input variables (operating parameters) within an operating range of the device under test for the one or more input variables.” Rivoir para 45 “However, the instrument 20 c may further provide curves 42 such as signal traces, spectra…” Signal traces are waveforms.)
a trained models store to store one or more trained neural networks, class variables and other configuration information for the trained neural network; and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
a communications store to store information received from the test automation application and from the machine learning system, the communications store configured to allow the test automation application and the machine learning system to access the information. (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error. … [0319] Examples are… Neural networks…” The classification algorithm runs on the test data and parameters, so it is allowed access to the information. Wherever the information and parameters are stored is the communications store.)
Rivoir doesn’t teach training.
However, Lan teaches a training data store, the training data store having substores for each value of a testing parameter for a device under test. (Lan abs “gathering state-action pair data from an expert policy; applying imitation learning to yield a cloned policy based on the gathered state-action pair data from the expert policy; and applying a reinforcement learning technique, wherein the reinforcement learning technique is initialized based on the cloned policy…” Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 11 “the cloned policy is in the form of a neural network, wherein the deepest hidden layer is convolutional in one dimension.”)
Rivoir teaches claim 14. A method comprising:
receiving and storing (Rivoir Fig. 3 below)
PNG
media_image1.png
334
794
media_image1.png
Greyscale
storing one or more trained neural networks and a configuration file in the structured data store; (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
receiving and storing reference parameters for the DUTs from the test automation application in a reference parameter store; (Rivoir para 27 “the test case generator is configured to randomly generate a plurality of test cases, wherein the test case generator is configured to randomly generate values for the one or more input variables (operating parameters) within an operating range of the device under test for the one or more input variables.”)
accessing the reference parameter store to retrieve the reference parameters and using the reference parameters to produce optimal tuning parameters for the DUTs; (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
storing the optimal tuning parameters in the structured data store; and (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…” if the combinations are predicted they are also stored somewhere because they are all inside of a computer system.)
allowing the test automation application to retrieve the optimal tuning parameters. (Rivoir para 42 “The instrument 20 a may derive stimulating settings 24 from the test cases 14 to provide a stimulus signal 24′ to the device under test.”)
Rivoir doesn’t teach training.
However, Lan teaches receiving and storing training data obtained from testing DUTs from a test automation application in a training store within a structured data store;
accessing the training store to retrieve the training data and using the training data to train one or more neural networks in a machine learning system; (Lan abs “gathering state-action pair data from an expert policy; applying imitation learning to yield a cloned policy based on the gathered state-action pair data from the expert policy; and applying a reinforcement learning technique, wherein the reinforcement learning technique is initialized based on the cloned policy…” Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 11 “the cloned policy is in the form of a neural network, wherein the deepest hidden layer is convolutional in one dimension.”)
Rivoir, Lan and the claims are using machine learning models on diagnostic data. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to train using Lan’s training method because “it can outperform a well-trained expert model with a higher tuning success rate and fewer adjustment steps which leads to shorter total tuning time.” Lan para 9.
Rivoir teaches claim 15. The method as claimed in claim 14, further comprising:
receiving and storing S-parameters obtained from the DUTs by the test automation application in the structured data store;
accessing the structured data store to retrieve the S-parameters and using the S-parameters to produce predicted performance parameters for the DUTs; and
storing the predicted performance parameters in the structured data store. (Rivoir para 50 “Calibration (or correction) exploits performance dependencies on settings or known conditions for adjustments to improve performance.” Rivoir para 64 “Pre-silicon, test engineers create highly parameterizable test methods, which give control over allowed combinations of stimuli 24′, DUT settings 34, operating conditions 26 a, 26 b, FW settings 32 (desired behavior, algorithm options). As part of this effort, FW engineers may implement a FW layer 22 that sets up the device as a function of desired behavior and equally parameterizable or randomizable algorithm options. These test methods also gather as much information about the DUT and its behavior as possible, including sensor data and on-chip status information.”)
Rivoir teaches claim 16. The method as claimed in claim 14, further comprising allowing the machine learning system to only access trained models developed in the machine learning system. (Rivoir only discloses trained models, para 318-319.)
Lan teaches claim 17. The method as claimed in claim 14, further comprising:
storing the training data, the reference parameters, and the trained neural networks
storing (Lan para 5 “Imitation learning learns how to make sequences of decisions in an environment, where the training signal comes from demonstrations.” Lan para 9 “the imitation and reinforcement learning based cavity filter tuning model of embodiments has been applied in a simulation environment and could tune cavity filters with more screws and satisfy both S11 and S21 parameters (return loss and insertion loss) and tuned both coupling and cross-coupling…”)
Lan doesn’t teach a global portion or prediction requests.
However, Huber teaches a global portion. (Huber fig. 1 global 20 and local 40. Huber claim 5 “global data used for a plurality of assembly lines and local data used at one of the plurality of assembly lines…”)
Rivoir, Lan, Huber and the claims all process data for manufacturing. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to have a global server that way “the differences between the data sets and the standardization data can be easily viewed using the difference editor.” Huber 2:15.
Lan doesn’t teach prediction requests.
However, Knaan teaches prediction requests. (Knaan fig. 3 and para 54 “trained machine learning model(s) 332 can receive input data 330 and one or more inference/prediction requests 340 (perhaps as part of input data 330) and responsively provide as an output one or more inferences and/or predictions 350.” The input data is in storage 330, so the inference requests are stored in input data storage 330.)
Lan, Knaan and the claims are using trained neural networks to make predictions. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to store prediction requests because the models are usually bottlenecks and requests shouldn’t be thrown away because they can’t immediately be run.
Rivoir teaches claim 18. The method as claimed in claim 14, further comprising storing the training data, reference parameters, trained neural networks, and optimal tuning parameters in a (Rivoir para 318-319 “Classification algorithms return a model that predicts which variable combinations will lead to an error…. [0319] Examples are… Neural networks…”)
Rivoir doesn’t have a global database.
However, Huber teaches parameters in a (Huber fig. 1 global 20 and local 40. Huber claim 5 “global data used for a plurality of assembly lines and local data used at one of the plurality of assembly lines…”)
Rivoir, Huber and the claims all process data for manufacturing. It would have been obvious to a person having ordinary skill in the art, at the time of filing, to have a global server that way “the differences between the data sets and the standardization data can be easily viewed using the difference editor.” Huber 2:15.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Austin Hicks whose telephone number is (571)270-3377. The examiner can normally be reached Monday - Thursday 8-4 PST.
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, Mariela Reyes can be reached at (571) 270-1006. 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.
/AUSTIN HICKS/ Primary Examiner, Art Unit 2142
Prosecution Insights