DETAILED ACTION
This action is in response to an amendment filed on April 20, 2026 for the application of Kwon et al., for a “Storage device predicting failure using machine learning and method of operating the same” filed on October 2, 2023. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 1-20 are pending in the application.
Claims 1, 3, 7-8, 11, and 15 have been amended.
Claims 1-20 are rejected under 35 USC § 101.
Claims 1-20 are rejected under 35 USC § 103.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claim(s) recite(s) mental processes-concepts performed in human mind.
As per claim 1, with the exception of the recitation of the limitation(s) “a storage device”, the limitations “identifying risk data from at least a portion of telemetry information, stored in a memory, based on a first criterion; inputting first data of a first attribute, from among the risk data, to a machine learning model; obtaining a first anomaly score output from the machine learning model; detecting whether an anomaly is present in the first attribute, based on a determination of whether the first anomaly score satisfies a second criterion; transmitting an alert, associated with the first attribute, to a host in response to the anomaly being detected, the alert enabling the host to determine whether the anomaly will cause the failure and to output feedback based on the determination; training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data; and controlling an operation of the storage device such that the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure.” can be performed by a human mind or with the aid of pen and paper. (MPEP 2106.04(a)(2)).
The limitations reciting steps to identify risk data, input first data into a model, obtaining a first anomaly score, detecting whether an anomaly is present based on the score, and adjusting the first attribute in which the anomaly is detected, merely require a person looking at collected data and thinking/analyzing the collected data by comparing the data and calculating a score. It is therefore a simple matter of a person looking at the data and thinking about the characteristics of the data to accomplish the above steps.
Step 2A. This judicial exception is not integrated into a practical application because the additional element(s) “a storage device” is/are directed to generic computer components recited at a high-level of generality such that they amount to nothing more than mere instructions to apply the exception using generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmitting an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity (MPEP 2106.05(g).
The limitation(s) of “controlling an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
Step 2B. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional element(s) “a storage device” does/do not provide significantly more than the recited judicial exception because the additional elements are mere instructions to implement an abstract idea or other exception on a computer and in this case generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmitting an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity. These limitations amount to transmitting data over a network and are well-understood, routine, conventional activity (MPEP 2106.05(d)).
The limitation(s) of “controlling an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
As for the limitations recited in claims 2-7, when considering each of the claims as a whole these additional elements do not integrate the exception into a practical application, using one or more of the considerations laid out by the Supreme Court and the Federal Circuit. The additional elements do not reflect an improvement in the functioning of a computer, or an improvement to other technology or technical field. The additional elements do not implement a judicial exception with, or use a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim. The additional element do not apply or use the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception.
As per claim 8, with the exception of the recitation of the limitation(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, the telemetry information obtained by monitoring the storage device, the controller including processing circuitry configured to store the telemetry information in at least one of the memory or the nonvolatile memory”, the limitations “identify risk data from at least a portion of the telemetry information, stored in at least one of the memory or the nonvolatile memory, based on a first criterion; input first data of a first attribute, from among the risk data, to a machine learning model; obtain a first anomaly score output from the machine learning model; detect whether an anomaly is present in the first attribute, based on a determination of whether the first anomaly score satisfies a second criterion; transmit an alert, associated with the first attribute, to a host in response to the anomaly being detected, the alert enabling the host to determine whether the anomaly will cause a failure and to output feedback based on the determination; and training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data; and controlling an operation of the storage device such that the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure.” can be performed by a human mind or with the aid of pen and paper. (MPEP 2106.04(a)(2)).
Step 2A. This judicial exception is not integrated into a practical application because the additional element(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, the telemetry information obtained by monitoring the storage device, the controller including processing circuitry configured to store the telemetry information in at least one of the memory or the nonvolatile memory” is/are directed to generic computer components recited at a high-level of generality such that they amount to nothing more than mere instructions to apply the exception using generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmit an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity (MPEP 2106.05(g).
The limitation(s) of “control an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
Step 2B. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional element(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, the telemetry information obtained by monitoring the storage device, the controller including processing circuitry configured to store the telemetry information in at least one of the memory or the nonvolatile memory” does/do not provide significantly more than the recited judicial exception because the additional elements are mere instructions to implement an abstract idea or other exception on a computer and in this case generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmit an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity. These limitations amount to transmitting data over a network and are well-understood, routine, conventional activity (MPEP 2106.05(d)).
The limitation(s) of “control an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
As per claim 15, with the exception of the recitation of the limitation(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, a controller comprising a memory configured to store telemetry information on the storage device the telemetry information obtained by monitoring the storage device, and processing circuitry configured to store risk data identified from the telemetry information, store a debug dump based on detection of an anomaly”, the limitations “identify risk data from at least a portion of telemetry information, stored in the memory, based on a first criterion, detect whether an anomaly is present in at least a portion of attributes, among the stored risk data, through a machine learning model trained using the identified risk data, transmit an alert, associated with an attribute in which the anomaly is detected, to a host in response to an anomaly being detected, the alert enabling the host to determine whether the anomaly will cause a failure and to output feedback based on the determination, train the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data, and control an operation of the storage device such that the attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure.” can be performed by a human mind or with the aid of pen and paper. (MPEP 2106.04(a)(2)).
Step 2A. This judicial exception is not integrated into a practical application because the additional element(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, and processing circuitry configured to store risk data identified from the telemetry information, store a debug dump based on detection of an anomaly” is/are directed to generic computer components recited at a high-level of generality such that they amount to nothing more than mere instructions to apply the exception using generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmit an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity (MPEP 2106.05(g).
The limitation(s) of “control an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model when the feedback received from the host is based on a determination that the anomaly will not cause the failure” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
Step 2B. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional element(s) “A storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device, and processing circuitry configured to store risk data identified from the telemetry information, store a debug dump based on detection of an anomaly” does/do not provide significantly more than the recited judicial exception because the additional elements are mere instructions to implement an abstract idea or other exception on a computer and in this case generic computer components (MPEP 2106.05(f)).
The limitation(s) of “transmit an alert” is/are mere data gathering recited at a high level of generality, and thus are insignificant extra-solution activity. These limitations amount to transmitting data over a network and are well-understood, routine, conventional activity (MPEP 2106.05(d)).
The limitation(s) of “control an operation of the storage device in response to receiving feedback” and “the first attribute in which the anomaly is detected is adjusted when the feedback received from the host is based on a determination that the anomaly will cause the failure” recite(s) steps that merely apply controlling operations, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
The limitation(s) of “wherein the machine learning model is configured to be trained on the risk data” and “training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data and to output anomaly scores based on the learned pattern of the data” recite(s) steps that merely apply a machine learning model for generic training, which represents merely adding the words “apply it”, or an equivalent, which are not indicative of an inventive concept (MPEP 2106.05(f)).
As per claims 9-14 and 16-20, please refer to analysis section for claims 2-7.
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 of this title, 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 set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-4, 8-12, and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Mcguinness et al. (U.S. PGPUB 20210342205) in view of Elmtalab et al. (U.S. PGPUB 20210200654).
As per claims 1 and 8, Mcguinness discloses a failure prediction method of predicting a failure of a storage device ([0014]), the failure prediction method comprising/a storage device comprising: a nonvolatile memory; and a controller comprising a memory configured to store telemetry information on the storage device ([0004]), the controller including processing circuitry configured to:
monitoring telemetry information of the storage device ([0013], “The management system 110 may include one or more computing devices that are configured to monitor the status of the storage device”) and ([0014], “telemetry data that is obtained from the storage device 134”);
storing the telemetry information in a memory ([0014], “The database 114 may be arranged to store one or more of: (i) telemetry data that is obtained from the storage device 134”);
identifying risk data from at least a portion of telemetry information, stored in a memory, based on a first criterion ([0014], “telemetry data that is obtained from the storage device 134 and/or (ii) a failure risk score that is calculated based on the telemetry data”);
inputting first data of a first attribute, from among the risk data, to a machine learning model (Fig. 1, “predictive failure model 120”); obtaining a first anomaly score output from the machine learning model ([0014], “The predictive failure model may receive, as input, telemetry data that is produced by a storage device 134 (which is part of the field system 120) and calculate, based on the telemetry data, a failure risk score that indicates the likelihood of the storage device 134 failing”);
detecting whether an anomaly is present in the first attribute, based on a determination of whether the first anomaly score satisfies a second criterion ([0025], “if the failure risk score is above a threshold (or within a first range), the risk score analyzer 116 may determine that the storage device 134 is at risk of failing”);
transmitting an alert, associated with the first attribute, to a host in response to the anomaly being detected ([0026], “As a result of executing the predictive failure model, the failure inference engine 122 may calculate a failure risk score for the storage device 134, which indicates the likelihood of the storage device 134 failing. After the failure risk score is calculated, the failure inference engine 122 may provide the failure risk score to the risk score analyzer 116.” and [0045]) and (Fig. 6), the alert enabling the host to determine whether the anomaly will cause the failure and to output feedback based on the determination (Fig. 2, elements 270, 280, and 210);
training the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data ([0027], “The predictive failure model may be trained by using a supervised learning algorithm and/or any other suitable type of training algorithm, such as an unsupervised anomaly detection algorithm that is arranged to detect anomalous behavior of storage devices by training on historic non-anomalous processed data of manufacturer-specific storage devices. The predictive failure model may be arranged to learn the behavior of a storage device telemetry and to detect major degradation patterns of the storage device that result in failure or imminent failure.”) and to output anomaly scores based on the learned pattern of the data ([0025]-[0026]); and
controlling an operation of the storage device ([0026], “Upon receiving the failure risk score for the storage device 134, the risk score analyzer 116 may execute a preemptive maintenance action for the storage device 134.” and [0045]) and (Fig. 6).
Mcguinness fails to explicitly disclose adjusting the first attribute in which the anomaly is detected.
Elmtalab of analogous art teaches controlling the operation of the storage device such that the first attribute in which the anomaly is detected is adjusted ([0028], “In response to determining the failure prediction 224, the NVDIMM 100 may execute remedial operations to lower the current temperatures of the controller 132 and/or the NV memory 130”) when the feedback received from the host is based on a determination that the anomaly will cause the failure ([0027], “The NVDIMM 100 can determine the failure prediction 224 when the temperature predictions exceed the temperature threshold 222.”).
All of the claimed elements were known in Mcguinness and Elmtalab and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art before the time of effective filing language to combine their storage failure detection methods. One would be motivated to make this combination since Elmtalab’s adjusting attributes is a mere example of Mcguinness’s preemptive maintenance action.
As per claims 2 and 9, Mcguinness discloses generating first variance data from the first data of the first attribute; inputting the first variance data to the machine learning model (Fig. 2); and obtaining a second anomaly score from the machine learning model based on the first variance data (Figs. 2 and 5); and determining whether an anomaly has occurred in the first attribute, in response to the second anomaly score satisfying a third criterion (Fig. 6) and ([0025]).
As per claims 3 and 11, Mcguinness discloses the monitoring the telemetry information ([0013], “The management system 110 may include one or more computing devices that are configured to monitor the status of the storage device”) includes monitoring the telemetry information during a first period ([0014], “telemetry data that is produced by a storage device 134”) to identify the at least the portion of the telemetry information as the risk data; and wherein the method further comprises storing the identified risk data in at least one of a nonvolatile memory or the memory ([0014] and [0034]).
As per claims 4 and 12, Mcguinness discloses the machine learning model is trained to output criteria modulated for at least one of the first criterion, the second criterion, and the third criterion (Fig. 3), based on at least one of the risk data or the feedback received from the host, and the failure prediction method further comprises at least one of identifying the risk data or determining whether an anomaly has occurred in the risk data, based on the modulated criteria (Figs. 2-6).
As per claim 10, Mcguinness discloses the controller is configured to infer a causal factor of the anomaly in response to the detection of the anomaly ([0028]-[0029]), and the alert transmitted to the host comprises the inferred causal factor ([0026], “As a result of executing the predictive failure model, the failure inference engine 122 may calculate a failure risk score for the storage device 134, which indicates the likelihood of the storage device 134 failing. After the failure risk score is calculated, the failure inference engine 122 may provide the failure risk score to the risk score analyzer 116.” and [0045]) and (Fig. 6).
As per claim 14, Mcguinness discloses the feedback comprises a control signal enabling the storage device to prevent the failure ([0029], “The preemptive maintenance action (that is executed based on the failure risk score) may include any suitable type of action for preventing a data loss that might result from a failure of the storage device 134.”), and the processing circuitry is further configured to control an operation of the storage device based on the control signal (Fig. 6, element 606, “EXECUTE THE PREEMPTIVE MAINTENANCE ACTION”).
Claims 5-7, 13 and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Mcguinness et al. (U.S. PGPUB 20210342205) in view of Elmtalab et al. (U.S. PGPUB 20210200654) and in further view of Karuppiah et al. (U.S. Patent No. 11422921).
As per claims 5 and 13, Mcguinness fails to explicitly disclose a debug feature.
Karuppiah of analogous art teaches enabling a debug feature, associated with the first attribute, in response to detecting the anomaly in the first attribute (col. 7, lines 14-35); determining whether the failure in the storage device has occurred, based on a failure criterion (col. 7, lines 29-35, “The debug HW 160 may be configured to receive a stream of FW events from the data storage device 120 and compare a FW event ID of each FW event to the ID of the triggering FW event. In response to detecting a match, the debug HW 160 may transmit a general purpose input/output (GPIO) signal to the cross feature hardware 170”); and storing a debug dump (col. 7, lines 23-29, “One or more memory devices of the debug HW 160 may comprise a controller read-only memory and/or a controller volatile memory (e.g., DRAM) in which the ID of a triggering firmware (FW) event may be stored along with firmware (FW) events written by the data storage device 120”), corresponding to the enabled debug feature in response to a determination that the failure has occurred in the storage device (Figs. 1B-1C).
All of the claimed elements were known in Mcguinness and Karuppiah and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art before the time of effective filing language to combine their storage failure detection methods. One would be motivated to make this combination for the purpose of improving validation of solid state drives (Karuppiah, col. 1, lines 19-25).
As per claim 6, Karuppiah discloses transmitting the stored debug dump to the host in response to receiving an information request for the debug dump from the host (col. 7, line 14 through col. 8, line 18).
As per claim 7, Mcguinness discloses the feedback comprises a control signal enabling the storage device to prevent the failure from occurring ([0029], “The preemptive maintenance action (that is executed based on the failure risk score) may include any suitable type of action for preventing a data loss that might result from a failure of the storage device 134.”), and the failure prediction method further comprises controlling the operation of the storage device based on the control signal (Fig. 6, element 606, “EXECUTE THE PREEMPTIVE MAINTENANCE ACTION”).
As per claim 15, Mcguinness discloses a storage device comprising:
a nonvolatile memory ([0015]); and a controller comprising a memory configured to store telemetry information on the storage device, the telemetry information obtained by monitoring the storage device ([0014], “The database 114 may be arranged to store one or more of: (i) telemetry data that is obtained from the storage device 134”), and
processing circuitry configured to store risk data identified from the telemetry information ([0014], “The database 114 may be arranged to store one or more of: (i) telemetry data that is obtained from the storage device 134 and/or (ii) a failure risk score that is calculated based on the telemetry data”), and
identify risk data from at least a portion of telemetry information, stored in the memory, based on a first criterion ([0014], “telemetry data that is obtained from the storage device 134 and/or (ii) a failure risk score that is calculated based on the telemetry data”),
detect whether an anomaly is present in at least a portion of attributes, among the stored risk data ([0025], “if the failure risk score is above a threshold (or within a first range), the risk score analyzer 116 may determine that the storage device 134 is at risk of failing”), through a machine learning model trained using the identified risk data ([0014], “The model provider 112 may include logic that is arranged to train a predictive failure model and provide the predictive failure model to the field system 120. The predictive failure model may receive, as input, telemetry data that is produced by a storage device 134 (which is part of the field system 120) and calculate, based on the telemetry data, a failure risk score that indicates the likelihood of the storage device 134 failing”),
transmit an alert, associated with an attribute in which the anomaly is detected, to a host in response to an anomaly being detected ([0026], “As a result of executing the predictive failure model, the failure inference engine 122 may calculate a failure risk score for the storage device 134, which indicates the likelihood of the storage device 134 failing. After the failure risk score is calculated, the failure inference engine 122 may provide the failure risk score to the risk score analyzer 116.” and [0045]) and (Fig. 6), the alert enabling the host to determine whether the anomaly will cause a failure and to output feedback based on the determination (Fig. 2, elements 270, 280, and 210); and
train the machine learning model based on the feedback received from the host such that the machine learning model is configured to learn a pattern of data from the risk data ([0027], “The predictive failure model may be trained by using a supervised learning algorithm and/or any other suitable type of training algorithm, such as an unsupervised anomaly detection algorithm that is arranged to detect anomalous behavior of storage devices by training on historic non-anomalous processed data of manufacturer-specific storage devices. The predictive failure model may be arranged to learn the behavior of a storage device telemetry and to detect major degradation patterns of the storage device that result in failure or imminent failure.”) and to output anomaly scores based on the learned pattern of the data ([0025]-[0026]); and
control an operation of the storage device ([0026], “Upon receiving the failure risk score for the storage device 134, the risk score analyzer 116 may execute a preemptive maintenance action for the storage device 134.” and [0045]) and (Fig. 6).
Mcguinness fails to explicitly disclose a debug dump.
Karuppiah of analogous art teaches store a debug dump based on detection of an anomaly (col. 7, lines 14-35).
All of the claimed elements were known in Mcguinness and Karuppiah and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art before the time of effective filing language to combine their storage failure detection methods. One would be motivated to make this combination for the purpose of improving validation of solid state drives (Karuppiah, col. 1, lines 19-25).
Mcguinness fails to explicitly disclose adjusting the first attribute in which the anomaly is detected.
Elmtalab of analogous art teaches controlling the operation of the storage device such that the first attribute in which the anomaly is detected is adjusted ([0028], “In response to determining the failure prediction 224, the NVDIMM 100 may execute remedial operations to lower the current temperatures of the controller 132 and/or the NV memory 130”) when the feedback received from the host is based on a determination that the anomaly will cause the failure ([0027], “The NVDIMM 100 can determine the failure prediction 224 when the temperature predictions exceed the temperature threshold 222.”).
All of the claimed elements were known in Mcguinness and Elmtalab and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art before the time of effective filing language to combine their storage failure detection methods. One would be motivated to make this combination since Elmtalab’s adjusting attributes is a mere example of Mcguinness’s preemptive maintenance action.
As per claim 16, please refer to analysis section for claim 5.
As per claim 17, Mcguinness discloses the nonvolatile memory comprises a risk data area, in which the risk data is stored ([0014], “The database 114 may be arranged to store one or more of: (i) telemetry data that is obtained from the storage device 134 and/or (ii) a failure risk score that is calculated based on the telemetry data”), and
Karuppiah discloses a debug dump area in which the debug dump corresponding to the enabled debug feature is stored (col. 7, lines 14-35).
As per claim 18, Mcguinness discloses the processing circuitry is configured to store the risk data in the risk data area of the nonvolatile memory ([0014]-[0015]) according to a preset cycle ([0034]).
As per claim 19, please refer to analysis section for claim 5.
As per claim 20, please refer to analysis section for claim 2.
Response to Arguments
Applicant’s amendments filed on April 20, 2026 have been fully considered but they are not persuasive.
With respect to the 35 U.S.C. 101 rejection, applicant's arguments have been fully considered but they are not persuasive.
Examiner would like to point out that as stated during an interview conducted on February 27, 2026, inclusion of the proposed amendment of “the training the machine learning model is performed using the telemetry information of a first period,
the controlling the operation of the storage device is performed using the telemetry information of a second period, and the first period is different from the second period” into independent claims, would overcome the 35 U.S.C. 101 rejections.
Applicant argues that Example 39 provides that training a neural network includes steps that are not practically performed in the human mind. The present claims include elements which should be considered eligible for similar reasons as provided in this example and “A comparison of the present claims to the above analysis, clearly shows that the present claims are more alike to the examples present in Example 39, as the claims do not require "specific mathematical calculations by referring to the mathematical calculations by name" and include "training the machine learning model" which the Kim memo expressly provides as examples of steps which cannot practically be performed in the human mind.”
The examiner respectfully disagrees and would like to point out that the limitations reciting steps to identify risk data, input first data into a model, obtaining a first anomaly score, detecting whether an anomaly is present based on the score, and adjusting the first attribute in which the anomaly is detected, merely require a person looking at collected data and thinking/analyzing the collected data by comparing the data and calculating a score. It is therefore a simple matter of a person looking at the data and thinking about the characteristics of the data to accomplish the above steps.
The direction to perform the mental process on a generic computer does not alter the fact that the claim recites a mental process. Further, a recitation at a high level of generality that a computer performs a step “wherein the machine learning model is configured to be trained on the risk data, to learn a pattern of data from the risk data, and to output anomaly scores based on the learned pattern of the data” does not provide more than a generic indication to use a computer as a tool to perform the action. Using a machine learning model with no indication of how the training/learning is performed is directed to a recitation at a high level of generality where uses a computer as a tool to perform a step that is readily performed mentally. If the process can be performed mentally even though the claimed procedures can be performed using a computer as a tool to perform them, the process is a mental process. See MPEP 2106.04(a)(2)(III)(C).
With respect to the 35 U.S.C. 103 rejection, applicant's arguments have been fully considered but they are not persuasive.
Applicant argues that “Mcguinness provides a predictive failure model which calculates a failure risk score of a storage device based on telemetry data. In Mcguinness, the predictive failure model is trained on a method for calculating the failure risk score using at least one of a training telemetry dataset 201, a training telemetry dataset 202, and a supervision dataset.
In contrast, the claims from the feature of Mcguinness in that, in the claims, risk data satisfying a specific criterion (e.g., a first criterion) is identified from telemetry information that is obtained by monitoring the storage device and stored (e.g., in real time), and trains a machine learning model with the identified data. Persons of ordinary skill in the art would recognize that the datasets of Mcguinness, which are "derived from testing performed at a storage device maintenance facility over the diagnosing and repairing storage devices" are different from the "telemetry information" of the claims, particularly in view of the claims as amended.”
The examiner respectfully disagrees and would like to point out to paragraph [0013], wherein Mcguinness discloses “The management system 110 may include one or more computing devices that are configured to monitor the status of the storage devices deployed at the field system 120”
Note paragraph [0014], wherein Mcguinness discloses “The database 114 may be arranged to store one or more of: (i) telemetry data that is obtained from the storage device 134”
Further note paragraph [0035], wherein Mcguinness discloses “the predictive failure model may be trained based on both anomalous and non-anomalous processed telemetry data across one or more storage devices.”
Monitoring storage devices by obtaining telemetry data and training a failure model with anomalous and non-anomalous processed telemetry data as disclosed by Mcguinness reads on the claimed limitations of claims 1, 8, and 15.
In response to applicant’s argument that there is no teaching, suggestion, or motivation to combine the Mcguinnes and Elmtalab references, the examiner recognizes that obviousness may be established by combining or modifying the teachings of the prior art to produce the claimed invention where there is some teaching, suggestion, or motivation to do so found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art. See In re Fine, 837 F.2d 1071, 5 USPQ2d 1596 (Fed. Cir. 1988), In re Jones, 958 F.2d 347, 21 USPQ2d 1941 (Fed. Cir. 1992), and KSR International Co. v. Teleflex, Inc., 550 U.S. 398, 82 USPQ2d 1385 (2007). In this case, both references disclose obtaining telemetry data and predicting failures based on the data (Mcguinness, Abstract) and (Elmtalab, [0009], “During run-time, the apparatus can obtain real-time temperature data of one or more circuits therein, such as for the controller and/or the FLASH memory”).
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Elmira Mehrmanesh whose telephone number is (571)272-5531. The examiner can normally be reached on M-F from 10-6.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Bryce Bonzo, can be reached at telephone number (571) 272-3655. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/Elmira Mehrmanesh/
Primary Examiner, Art Unit 2113
Prosecution Insights