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 .
Detailed Action
This action is in response to the claims filed 1/24/2024:
Claims 1 – 20 are pending.
Claims 1, 11, and 14 are independent.
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 10-13 and 20 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.
Regarding claim 10, "the sampling ratio" lacks antecedent basis. Claim 9 introduces a "sampling rate", however, the instant specification explicitly distinguishes a sampling rate from an extraction ratio such that it's unclear what is intended by the claim. This is further complicated by the fact that claim 10 recites "change the sampling rate" and simultaneously "maintain the sampling ratio". In the interest of further examination the claim is interpreted as "maintain the sampling rate based on the pass result data predicted as being the failure by the AI predictive model and determined as being normal as the result of the actual test not being detected".
Regarding claim 11, "AI" is an undefined abbreviation. It is unclear whether "AI" refers to artificial intelligence generally, a particular class of machine-learning algorithms, or another technology.
Regarding claim 20, "the controlling" lacks antecedent basis. "The method of claim 14, further comprising:" is recommended.
Claims 12-13 are rejected with respect to their dependence on rejected claim 11.
Claim Rejections - 35 USC § 101
101 Rejection
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 1-20 are rejected under 35 USC § 101 because the claimed invention is directed to non-statutory subject matter.
Regarding Claim 1: Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 1 is directed to an electronic device, which is directed to a product, one of the statutory categories.
Step 2A Prong One Analysis: Claim 1 under its broadest reasonable interpretation is a series of mental processes. For example, but for the generic computer components language, the above limitations in the context of this claim encompass machine learning processing, including the following:
generate an artificial intelligence (AI) predictive model, based on component input data; (observation, evaluation, and judgement),
update the AI predictive model, based on at least one of the component input data, the fail result data, and the pass result data (observation, evaluation, and judgement)
Therefore, claim 1 recites an abstract idea which is a judicial exception.
Step 2A Prong Two Analysis: Claim 1 recites additional elements “a communication module comprising communication circuitry; a memory; at least one processor comprising processing circuitry operatively connected to the communication module or the memory, wherein at least one processor is configured to”. However, these additional features are computer components recited at a high-level of generality, such that they amount to no more than mere instructions to apply the judicial exception using a generic computer component. An additional element that merely recites the words “apply it” (or an equivalent) with the judicial exception, or merely includes instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, does not integrate the judicial exception into a practical application (See MPEP 2106.05(f)). Claim 1 also recites additional elements “obtain fail result data according to the AI predictive model; obtain pass result data according to the fail result data;” which amounts to gathering and outputting data, which is insignificant extra-solution activity (See MPEP 2106.05(g)). Therefore, claim 1 is directed to a judicial exception.
Step 2B Analysis: Claim 1 does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the lack of integration of the abstract idea into a practical application, the additional elements recited in claim 1 amount to no more than mere instructions to apply the judicial exception using a generic computer component and insignificant extra-solution activity. The gathering and outputting of data is considered well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i)).
For the reasons above, claim 1 is rejected as being directed to non-patentable subject matter under §101. This rejection applies equally to independent claim 15, which recites a method, as well as to dependent claims 2-10 and 16-20. The additional limitations of the dependent claims are addressed briefly below:
Dependent claims 2 and 15 recite additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “transmit at least one of the component input data, the fail result data, or the pass result data to a server through the communication module; and receive the AI predictive model from the server” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i))
Dependent claim 3 recites additional observation, evaluation, and judgement “update the AI predictive model based on the component input data, the fail result data, or the pass result data being obtained” as well as additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “obtain fail result data and pass result data predicted by the updated AI predictive model” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i))
Dependent claim 4 recites additional observation, evaluation, and judgement “wherein the fail result data includes data predicted as being normal or faulty by the AI predictive model and determined as being faulty as a result of an actual test, and wherein the pass result data is data determined as being normal or faulty by the AI predictive model and determined as being normal based on the actual test.”
Dependent claim 5 recites additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “wherein at least one processor is configured to obtain the pass result data, based on a number of pieces of the fail result data.” Which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i))
Dependent claims 6 and 16 recite additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “obtain the pass result data according to a set ratio based on the number of pieces of the fail result data being less than or equal to a threshold value.” And “obtain the pass result data according to the changed extraction ratio based on the number of pieces of the fail result data exceeding the threshold value” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i)). Dependent claim 6 also recites additional observation, evaluation, and judgement “change an extraction ratio of the pass result data”
Dependent claims 7 and 17 recite additional observation, evaluation, and judgement “to update the AI predictive model, based on fail result data predicted as being normal by the AI predictive model and determined as a failure as a result of an actual test.”
Dependent claims 8 and 18 recite additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “to obtain the pass result data according to a configured sampling rate based on fail result data predicted as being a failure by the AI predictive model and determined as a failure as a result of an actual test” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i))
Dependent claims 9 and 19 recite additional observation, evaluation, and judgement “to control a sampling rate at which the fail result data and the pass result data are inspected, based on pass result data predicted as being a failure by the AI predictive model and determined as being normal as a result of an actual test”
Dependent claims 10 and 20 recite additional observation, evaluation, and judgement “change the sampling rate based on the pass result data predicted as being the failure by the AI predictive model and determined as being normal as the result of the actual test being detected; and maintain the sampling ratio based on the pass result data predicted as being the failure by the AI predictive model and determined as being normal as the result of the actual test not being detected.”
Regarding Claim 11: Claim 11 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 11 is directed to an AI prediction system comprising: an electronic device comprising circuitry, which is directed to a product, one of the statutory categories.
Step 2A Prong One Analysis: Claim 11 under its broadest reasonable interpretation is a series of mental processes. For example, but for the generic computer components language, the above limitations in the context of this claim encompass machine learning processing, including the following:
generate or update the AI predictive model, based on at least one of the component input data, the fail result data, or the pass result data; (observation, evaluation, and judgement),
Therefore, claim 11 recites an abstract idea which is a judicial exception.
Step 2A Prong Two Analysis: Claim 11 recites additional elements “a communication module comprising communication circuitry; a memory; at least one processor comprising processing circuitry operatively connected to the communication module or the memory, wherein at least one processor is configured to”. However, these additional features are computer components recited at a high-level of generality, such that they amount to no more than mere instructions to apply the judicial exception using a generic computer component. An additional element that merely recites the words “apply it” (or an equivalent) with the judicial exception, or merely includes instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, does not integrate the judicial exception into a practical application (See MPEP 2106.05(f)). Claim 11 also recites additional elements “obtain fail result data according to an AI predictive model, obtain pass result data according to the fail result data, and transmit at least one of component input data, the fail result data, or the pass result data to a server;” and “transmit the AI predictive model to the electronic device” which amounts to gathering and outputting data, which is insignificant extra-solution activity (See MPEP 2106.05(g)). Therefore, claim 11 is directed to a judicial exception.
Step 2B Analysis: Claim 11 does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the lack of integration of the abstract idea into a practical application, the additional elements recited in claim 11 amount to no more than mere instructions to apply the judicial exception using a generic computer component and insignificant extra-solution activity. The gathering and outputting of data is considered well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i)).
For the reasons above, claim 11 is rejected as being directed to non-patentable subject matter under §101. This rejection applies equally to dependent claims 12-13. The additional limitations of the dependent claims are addressed briefly below:
Dependent claim 12 recites additional observation, evaluation, and judgement “update the AI predictive model whenever the component input data, the fail result data, or the pass result data is obtained from the electronic device, and transmit the AI predictive model to the electronic device” as well as additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “to obtain fail result data and pass result data predicted by the AI predictive model received from the server” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i))
Dependent claim 13 recites additional insignificant extra-solution activity of gathering and outputting data (See MPEP 2106.05(g)) “obtain the pass result data according to a set ratio based on the number of pieces of the fail result data being less than or equal to a threshold value” which is well-understood, routine, and conventional in the art (See MPEP 2106.05(d)(II)(i)) as well as additional observation, evaluation, and judgement “change an extraction ratio of the pass result data and obtain the pass result data according to the changed extraction ratio based on the number of pieces of the fail result data exceeding the threshold value”
Therefore, when considering the elements separately and in combination, they do not add significantly more to the inventive concept. Accordingly, claims 1-20 are rejected under 35 U.S.C. § 101.
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 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)(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.
Claims 1, 3-10, 14, and 16-20 are rejected under U.S.C. §102(a)(1) as being anticipated by Chang (US20080250265A1).
PNG
media_image1.png
402
678
media_image1.png
Greyscale
FIG. 4 of Chang
Regarding claim 1, Chang teaches An electronic device comprising: a communication module comprising communication circuitry;( [¶0034] "Network adapters may also be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modem and Ethernet cards are just a few of the currently available types of network adapters.")
a memory; and at least one processor comprising processing circuitry operatively connected to the communication module or the memory,( [¶0033] "A data processing system suitable for storing and/or executing program code may include at least one processor coupled directly or indirectly to memory elements through a system bus. The memory elements can include local memory employed during actual execution of the program code, bulk storage, and cache memories which provide temporary storage of at least some program code to reduce the number of times code is retrieved from bulk storage during execution. Input/output or I/O devices (including but not limited to keyboards, displays, pointing devices, etc.) may be coupled to the system either directly or through intervening I/O controllers.")
wherein at least one processor is configured to: generate an artificial intelligence (AI) predictive model, based on component input data;( [Abstract] "to continuously monitor and collect operation states of different system components. An analysis subsystem is configured to build classification models to perform on-line failure predictions" [¶0013] "An analysis subsystem is configured to build classification models to perform on-line failure predictions. A failure prevention subsystem is configured to take preventive actions on failing components based on failure warnings generated by the analysis subsystem." [¶0074] "a classifier and classification method are illustratively described. A wide array of statistical or machine learning models may be used, which can also be efficiently updated, such as decision trees, Gaussian mixture models, or support vector machines" operation states interpreted as component input data. Statistical and machine learning models interpreted as synonymous with AI predictive model.)
obtain fail result data according to the AI predictive model;( [¶0037] "for each failure type, an analysis component or module 112 maintains a failure prediction model 114 that can classify all sample values associated with the failure type into three states: normal, pre-failure and failure.")
obtain pass result data according to the fail result data; and( [¶0037] "for each failure type, an analysis component or module 112 maintains a failure prediction model 114 that can classify all sample values associated with the failure type into three states: normal, pre-failure and failure." [¶0096] "If the object state is classified as normal later or the inspection times out, the system considers the prediction model issues a false-alarm" "Normal" classification interpreted as synonymous with obtained pass result data. Obtaining normal classification as a result of false-alarm interpreted as obtaining pass result data according to the fail result data)
update the AI predictive model, based on at least one of the component input data, the fail result data, and the pass result data.( [¶0039] "Upon receiving a failure warning, an inspection component or action module 118 can take proper preventive actions (e.g., isolating “bad” operators, migrating “good” operators, creating replacement operators). The inspection component 118 can provide feedback 120 to the analysis component 112. Based on the feedback information, the prediction models 114 can continuously evolve themselves to adapt to dynamic stream environments.").
Regarding claim 3, Chang teaches The electronic device of claim 1, wherein at least one processor is configured to: update the AI predictive model based on the component input data, the fail result data, or the pass result data being obtained; and(Chang [¶0039] "Upon receiving a failure warning, an inspection component or action module 118 can take proper preventive actions (e.g., isolating “bad” operators, migrating “good” operators, creating replacement operators). The inspection component 118 can provide feedback 120 to the analysis component 112. Based on the feedback information, the prediction models 114 can continuously evolve themselves to adapt to dynamic stream environments." [¶0088] " the feedback from the inspection component provides the prediction model with the correct label (i.e., normal or failure) about a historical measurement sample <m1,t . . . , mj,t> that can be used as training data for the decision tree ensembles")
obtain fail result data and pass result data predicted by the updated AI predictive model. (Chang [¶0087] "it can continuously update its prediction model to make more accurate and comprehensive failure predictions (e.g., make proper predictions under different workloads)").
Regarding claim 4, Chang teaches The electronic device of claim 1, wherein the fail result data includes data predicted as being normal or faulty by the AI predictive model and determined as being faulty as a result of an actual test, and (Chang [¶0050] "Let Nfp denote the false positive predictions where the prediction model sends alarms during the normal state, and Ntn denote true negative cases. On the other hand, predictive failure management also incurs extra resource cost because of the preventive actions" [¶0087] "If the primal decision tree makes a positive prediction (e.g., raising alarm) and the candidate decision tree makes a negative prediction (e.g., normal), a false-positive feedback to the primal decision tree is converted to a true-positive feedback to the candidate decision tree. " Chang explicitly distinguishes positive predictions/alarms and negative normal predictions and tracks both false negatives and true positive. A false negative is the case where the model did not warn but failure occurred; a true positive is the case where the model warned and failure occurred. That corresponds to fail data that may have been predicted normal or faulty)
wherein the pass result data is data determined as being normal or faulty by the AI predictive model and determined as being normal based on the actual test.(Chang [¶0050] "Let Nfp denote the false positive predictions where the prediction model sends alarms during the normal state, and Ntn denote true negative cases. On the other hand, predictive failure management also incurs extra resource cost because of the preventive actions" [¶0087] "If the primal decision tree makes a positive prediction (e.g., raising alarm) and the candidate decision tree makes a negative prediction (e.g., normal), a false-positive feedback to the primal decision tree is converted to a true-positive feedback to the candidate decision tree. " Chang explicitly distinguishes positive predictions/alarms and negative normal predictions and tracks both false negatives and true positive. A false negative is the case where the model did not warn but failure occurred; a true positive is the case where the model warned and failure occurred. That corresponds to fail data that may have been predicted normal or faulty).
Regarding claim 5, Chang teaches The electronic device of claim 1, wherein at least one processor is configured to obtain the pass result data, based on a number of pieces of the fail result data.(Chang [¶0096] " If the object state is classified as normal later or the inspection times out, the system considers the prediction model issues a false-alarm.").
Regarding claim 6, Chang teaches The electronic device of claim 5, wherein at least one processor is configured to: obtain the pass result data according to a set ratio based on the number of pieces of the fail result data being less than or equal to a threshold value; and (Chang [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] he analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." [¶0027] "sampling rate is dynamically adjusted based on the failure prediction results. For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost. When the prediction model raises an alarm on the object, the sampling rate is increased to collect more precise information on abnormal objects" Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous. The threshold value in Chang is 0. When the fail result data exceeds 0 the sampling ratio is changed.)
change an extraction ratio of the pass result data and obtain the pass result data according to the changed extraction ratio based on the number of pieces of the fail result data exceeding the threshold value. (Chang [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] he analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." [¶0027] "sampling rate is dynamically adjusted based on the failure prediction results. For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost. When the prediction model raises an alarm on the object, the sampling rate is increased to collect more precise information on abnormal objects" Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous. The threshold value in Chang is 0. When the fail result data exceeds 0 the sampling ratio is changed.).
Regarding claim 7, Chang teaches The electronic device of claim 1, wherein at least one processor is configured to update the AI predictive model, (Chang [¶0039] "Upon receiving a failure warning, an inspection component or action module 118 can take proper preventive actions (e.g., isolating “bad” operators, migrating “good” operators, creating replacement operators). The inspection component 118 can provide feedback 120 to the analysis component 112. Based on the feedback information, the prediction models 114 can continuously evolve themselves to adapt to dynamic stream environments." [¶0050] "Let Nfp denote the false positive predictions where the prediction model sends alarms during the normal state, and Ntn denote true negative cases. On the other hand, predictive failure management also incurs extra resource cost because of the preventive actions")
based on fail result data predicted as being normal by the AI predictive model and determined as a failure as a result of an actual test.(Chang [¶0087] "If the primal decision tree makes a positive prediction (e.g., raising alarm) and the candidate decision tree makes a negative prediction (e.g., normal), a false-positive feedback to the primal decision tree is converted to a true-positive feedback to the candidate decision tree. " Chang explicitly discloses that when the model misses a failure, the analysis component retrieves missed fine-grained logs, diagnoses the failure, labels training streams, creates decision trees, and updates the model so future failures are not missed).
Regarding claim 8, Chang teaches The electronic device of claim 1, wherein at least one processor is configured to obtain the pass result data according to a configured sampling rate (Chang [¶0014] "continuously monitoring and collecting" [¶0027] " For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost" [¶0059] "For each metric, its dynamic values are sampled periodically at a certain rate 308 to form a time series" [¶0039] "the prediction models 114 can continuously evolve themselves to adapt to dynamic stream environments." Chang explicitly uses a configured sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous.)
based on fail result data predicted as being a failure by the AI predictive model and determined as a failure as a result of an actual test.(Chang [¶0050] "Let Nfp denote the false positive predictions where the prediction model sends alarms during the normal state, and Ntn denote true negative cases. On the other hand, predictive failure management also incurs extra resource cost because of the preventive actions" [¶0087] "If the primal decision tree makes a positive prediction (e.g., raising alarm) and the candidate decision tree makes a negative prediction (e.g., normal), a false-positive feedback to the primal decision tree is converted to a true-positive feedback to the candidate decision tree. " Chang explicitly distinguishes positive predictions/alarms and negative normal predictions and tracks both false negatives and true positive. A false negative is the case where the model did not warn but failure occurred; a true positive is the case where the model warned and failure occurred. That corresponds to fail data that may have been predicted normal or faulty).
Regarding claim 9, Chang teaches The electronic device of claim 1, wherein at least one processor is configured to control a sampling rate at which the fail result data and the pass result data are inspected, (Chang [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] he analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous.)
based on pass result data predicted as being a failure by the AI predictive model and determined as being normal as a result of an actual test.(Chang [¶0050] "Let Nfp denote the false positive predictions where the prediction model sends alarms during the normal state, and Ntn denote true negative cases. On the other hand, predictive failure management also incurs extra resource cost because of the preventive actions" [¶0087] "If the primal decision tree makes a positive prediction (e.g., raising alarm) and the candidate decision tree makes a negative prediction (e.g., normal), a false-positive feedback to the primal decision tree is converted to a true-positive feedback to the candidate decision tree. " Chang explicitly distinguishes positive predictions/alarms and negative normal predictions and tracks both false negatives and true positive. A false negative is the case where the model did not warn but failure occurred; a true positive is the case where the model warned and failure occurred. That corresponds to fail data that may have been predicted normal or faulty).
Regarding claim 10, Chang teaches The electronic device of claim 9, wherein at least one processor is configured to: change the sampling rate based on the pass result data predicted as being the failure by the AI predictive model and determined as being normal as the result of the actual test being detected; and(Chang [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] he analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." [¶0027] "sampling rate is dynamically adjusted based on the failure prediction results. For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost. When the prediction model raises an alarm on the object, the sampling rate is increased to collect more precise information on abnormal objects" Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous.)
maintain the sampling ratio based on the pass result data predicted as being the failure by the AI predictive model and determined as being normal as the result of the actual test not being detected.(Chang [¶0095] "Based on the detailed measurement information, the prediction model continues to evaluate the object state as long as the object stays in the pre-failure state. To avoid infinite inspection on the object that is mistakenly classified as pre-failure by the prediction model, we set an upper-bound on the inspection time based on the past observations (e.g., maximum observed time-to-failure interval). The inspection will be automatically released when the inspection times out. If a failure does happen later, as predicted, the object transfers into the repair mode.").
Regarding claims 14 and 16-20, claims 14 and 16-20 are directed towards the method performed by the device of claims 1 and 6-10, respectively. Therefore, the rejections applied to claims 1 and 6-10 also apply to claims 14 and 16-20.
Claim 11 is rejected under U.S.C. §102(a)(1) as being anticipated by Ouyang (US20220091576A1).
Regarding claim 11, Ouyang teaches An AI prediction system comprising: ( [¶0016] "the approach can be described as a distributed AI quality inspection system")
an electronic device comprising circuitry configured to: ( [¶0020] "Quality inspection environment includes network 101, client computing device 102 and server 110.")
obtain fail result data according to an AI predictive model, ( [¶0036] "(15) Model controller collects the meta data about the product to be inspected and selects the model with highest score (method of scoring was described in model data component 211) to find out defect(s)" Defects interpreted as fail result data.)
obtain pass result data according to the fail result data, ( [¶0036] "if there is no defect then deploys the next model using model controller" [¶0029] "ranking the models and assigning a score to each model is based on the following equation: score=Σi=1 ndtlb/ntlb*a+(Σi=1 ndtl/ntl+Σi=1 ndlb/nlb+Σi=1 ndtb/ntb)*b+(Σi=1 ndt/nt+Σi=1 ndl/nl+Σi=1 ndb/nb)*c where dtlb—number of defects in products with the same type, same line and same batch, ntlb—number of inspected products with the same type" Ouyang explicitly obtains no-defect/pass information and uses defect counts with inspected product counts to drive model selection)
and transmit at least one of component input data, the fail result data, or the pass result data to a server, and( [¶0021] ", network 101 can be any combination of connections and protocols that can support communications between server 110, client computing device 102 and other computing devices (not shown) within quality inspection environment" [0032] " device controller component 213 of the present invention provides the capability of communicating and controlling a computer vision system. For example, device controller component 213 can instruct a computer vision system to begin inspecting product for defects based on the selected model. Device controller component 213 can also receive defect data related to the inspection of the product from the computer vision system." See FIG. 2)
the server is configured to: generate or update the AI predictive model, based on at least one of the component input data, the fail result data, or the pass result data, ( [¶0028] "model data component 211 of the present invention provides the capability of creating, ranking, storing and updating models associated with products from a manufacturing line." [¶0030] "The whole scoring calculation is done outside edge device (on an edge server) and works with the edge service to deliver the correct model for quality inspection. Any servers can be used to run the calculation/algorithm. However, it is recommended (i.e., best method) to use edge service since it provides a centralized management across all edge devices." See FIG. 2)
and transmit the AI predictive model to the electronic device.( [¶0016] " the approach can be described as a distributed AI quality inspection system. Distributed in this context means that the approach can manage multiple devices and manage multiple models across all edge devices. Edge service provides flexibility to deploy models and updates versions remotely based on inspected product type." [¶0030] "The whole scoring calculation is done outside edge device (on an edge server) and works with the edge service to deliver the correct model for quality inspection. Any servers can be used to run the calculation/algorithm. However, it is recommended (i.e., best method) to use edge service since it provides a centralized management across all edge devices." See FIG. 2).
Claim Rejections - 35 USC § 103
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 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 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 2 and 15 are rejected under U.S.C. §103 as being unpatentable over the combination of Chang and Ouyang.
Regarding claim 2, Chang teaches The electronic device of claim 1.
However, Chang doesn't explicitly teach wherein at least one processor is configured to: transmit at least one of the component input data, the fail result data, or the pass result data to a server through the communication module; and
receive the AI predictive model from the server.
Ouyang, in the same field of endeavor, teaches at least one processor is configured to: transmit at least one of the component input data, the fail result data, or the pass result data to a server through the communication module; ( [¶0021] ", network 101 can be any combination of connections and protocols that can support communications between server 110, client computing device 102 and other computing devices (not shown) within quality inspection environment" [0032] " device controller component 213 of the present invention provides the capability of communicating and controlling a computer vision system. For example, device controller component 213 can instruct a computer vision system to begin inspecting product for defects based on the selected model. Device controller component 213 can also receive defect data related to the inspection of the product from the computer vision system." See FIG. 2)
receive the AI predictive model from the server.( [¶0016] " the approach can be described as a distributed AI quality inspection system. Distributed in this context means that the approach can manage multiple devices and manage multiple models across all edge devices. Edge service provides flexibility to deploy models and updates versions remotely based on inspected product type." [¶0030] "The whole scoring calculation is done outside edge device (on an edge server) and works with the edge service to deliver the correct model for quality inspection. Any servers can be used to run the calculation/algorithm. However, it is recommended (i.e., best method) to use edge service since it provides a centralized management across all edge devices." See FIG. 2).
Chang as well as Ouyang are directed towards failure detection with machine learning models. Therefore, Chang as well as Ouyang are analogous art in the same field of endeavor. It would have been obvious before the effective filing date of the claimed invention to combine the teachings of Chang with the teachings of Ouyang by using the server in Ouyang for model serving of the continuously updated model in Chang. Ouyang provides as additional motivation for combination ([¶0002] “edge computing can improve response times and save bandwidth in an environment. Using edge computing with AI, it is a way to distribute AI and get result in real-time. Simulating and running models requires a lot of compute resources, thus, EDGE computing has an advantage over other computing system with regards to AI.”). This motivation for combination also applies to the remaining claims which depend on this combination.
Regarding claim 15, claim 15 is directed towards the method performed by the device of claim 2. Therefore, the rejection applied to claim 2 also applies to claim 15.
Claims 12 and 13 are rejected under U.S.C. §103 as being unpatentable over the combination of Ouyang and Chang.
Regarding claim 12, Ouyang teaches and transmit the AI predictive model to the electronic device, and (Ouyang [¶0016] " the approach can be described as a distributed AI quality inspection system. Distributed in this context means that the approach can manage multiple devices and manage multiple models across all edge devices. Edge service provides flexibility to deploy models and updates versions remotely based on inspected product type." [¶0030] "The whole scoring calculation is done outside edge device (on an edge server) and works with the edge service to deliver the correct model for quality inspection. Any servers can be used to run the calculation/algorithm. However, it is recommended (i.e., best method) to use edge service since it provides a centralized management across all edge devices." See FIG. 2)
wherein the electronic device configured to obtain fail result data and pass result data predicted by the AI predictive model received from the server.(Ouyang [¶0036] " Edge device invokes inspection, (18) Dashboard uses REST API of model to get results and if there is no defect then deploys the next model using model controller; repeat (16)-(18) and (19) If any defect was found then quality inspector receives alert and reviews results (i.e., the faulty product is sent to rework/repair)." [¶0039] "analysis component 214 can instruct (via device controller component 213) a computer vision system to begin inspecting the product (i.e. product_A). Product_A has three areas for inspection. Only one area passed inspection (i.e., Product_A_point3) but the entire product is considered defective. It is noted that the defect threshold can be adjusted by the user and/or AI system.").
However, Ouyang doesn't explicitly teach The AI prediction system of claim 11, wherein the server is configured to: update the AI predictive model whenever the component input data, the fail result data, or the pass result data is obtained from the electronic device.
Chang, in the same field of endeavor, teaches The AI prediction system of claim 11, wherein the server is configured to: update the AI predictive model whenever the component input data, the fail result data, or the pass result data is obtained from the electronic device, ( [¶0039] "Upon receiving a failure warning, an inspection component or action module 118 can take proper preventive actions (e.g., isolating “bad” operators, migrating “good” operators, creating replacement operators). The inspection component 118 can provide feedback 120 to the analysis component 112. Based on the feedback information, the prediction models 114 can continuously evolve themselves to adapt to dynamic stream environments.").
Chang as well as Ouyang are directed towards failure detection with machine learning models. Therefore, Chang as well as Ouyang are analogous art in the same field of endeavor. It would have been obvious before the effective filing date of the claimed invention to combine the teachings of Chang with the teachings of Ouyang by using the server in Ouyang for model serving of the continuously updated model in Chang. Ouyang provides as additional motivation for combination ([¶0002] “edge computing can improve response times and save bandwidth in an environment. Using edge computing with AI, it is a way to distribute AI and get result in real-time. Simulating and running models requires a lot of compute resources, thus, EDGE computing has an advantage over other computing system with regards to AI.”). This motivation for combination also applies to the remaining claims which depend on this combination.
Regarding claim 13, Ouyang teaches The AI prediction system of claim 11.
However, Ouyang doesn't explicitly teach, wherein the electronic device is configured to: obtain the pass result data according to a set ratio based on the number of pieces of the fail result data being less than or equal to a threshold value, and
change an extraction ratio of the pass result data and obtain the pass result data according to the changed extraction ratio based on the number of pieces of the fail result data exceeding the threshold value.
Chang, in the same field of endeavor, teaches The AI prediction system of claim 11, wherein the electronic device is configured to: obtain the pass result data according to a set ratio based on the number of pieces of the fail result data being less than or equal to a threshold value, and ( [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] the analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." [¶0027] "sampling rate is dynamically adjusted based on the failure prediction results. For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost. When the prediction model raises an alarm on the object, the sampling rate is increased to collect more precise information on abnormal objects" Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous. The threshold value in Chang is 0. When the fail result data exceeds 0 the sampling ratio is changed.)
change an extraction ratio of the pass result data and obtain the pass result data according to the changed extraction ratio based on the number of pieces of the fail result data exceeding the threshold value. ( [¶0062] "The basic idea is to use a low sampling rate when the analysis result is normal and switch to a high sampling rate during the analysis result is abnormal. […] the analysis component 112 makes the adjustment of the sampling rate 308 to sampling block 302. When the prediction model raises a failure alarm on the object, the sampling rate is increased to collect more precise measurement on the suspicious object, which allows the prediction model to make more accurate state classifications and time-to-failure estimations." [¶0027] "sampling rate is dynamically adjusted based on the failure prediction results. For each monitored object whose state is classified as normal, a low sampling rate is used to reduce resource cost. When the prediction model raises an alarm on the object, the sampling rate is increased to collect more precise information on abnormal objects" Chang explicitly using a configured, dynamic sampling rate for ingesting data used for prediction and subsequent fine-tuning. Chang is explicit that this process is continuous. The threshold value in Chang is 0. When the fail result data exceeds 0 the sampling ratio is changed.).
Chang as well as Ouyang are directed towards failure detection with machine learning models. Therefore, Chang as well as Ouyang are analogous art in the same field of endeavor. It would have been obvious before the effective filing date of the claimed invention to combine the teachings of Chang with the teachings of Ouyang by using the server in Ouyang for model serving of the continuously updated model in Chang. Ouyang provides as additional motivation for combination ([¶0002] “edge computing can improve response times and save bandwidth in an environment. Using edge computing with AI, it is a way to distribute AI and get result in real-time. Simulating and running models requires a lot of compute resources, thus, EDGE computing has an advantage over other computing system with regards to AI.”). This motivation for combination also applies to the remaining claims which depend on this combination.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Kaul (“AI-Driven Fault Detection and Self-Healing Mechanisms in Microservices Architectures for Distributed Cloud Environments”, 2020) is directed towards using machine learning for failure detection in a distributed cloud environment, and self-healing through continuous learning.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SIDNEY VINCENT BOSTWICK whose telephone number is (571)272-4720. The examiner can normally be reached M-F 7:30am-5:00pm 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, Miranda Huang can be reached on (571)270-7092. 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.
/SIDNEY VINCENT BOSTWICK/Examiner, Art Unit 2124
Prosecution Insights