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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on September 03, 2026 has been entered.
Notice to Applicant
The following is a Non-Final Office Action for Application Serial Number: 18/317,801, filed on May 15, 2023. In response to Examiner's Final Office Action dated June 03, 2026, Applicant on September 03, 2026, amended claims 1, 4, 5, 7, 8, 12 and 14 and cancelled claims 3, 6 and 11. Claims 1, 4, 5, 7, 8, 10, and 12-15 are pending in this application and have been rejected below.
Response to Amendment
Applicant's amendments are acknowledged.
Regarding the 35 U.S.C. 101 rejection, Applicants arguments and amendments have been considered but are insufficient to overcome the rejection.
The 35 U.S.C. § 103 rejection of claims 1, 4, 5, 7, 8, 10 and 12-15 are hereby amended pursuant Applicant’s amendments to claim 1.
Response to Arguments
Applicant's Arguments/Remarks filed September 03, 2026 (hereinafter Applicant Remarks) have been fully considered but are not persuasive. Applicant’s Remarks will be addressed herein below in the order in which they appear in the response filed September 03, 2026.
Regarding the 35 U.S.C. 101 rejection, Applicant states none of the above features (see p. 6-7, Applicant Remarks) recite a fundamental economic principle or practice, a commercial or legal interaction, or managing personal behavior or relationships or interactions between people-the three enumerated sub-groupings of "certain methods of organizing human activity." MPEP § 2106.04(a)(2) states that those sub-groupings are "not to be expanded beyond ... except in rare circumstances." Applicant submits that training a machine learning model to predict error conditions of a software robot using time-stamped workflow event data and confirmed user input is a technical process directed to improving error prediction for that robot, not a commercial interaction or business process.
The Office further noted that "[t]he claim currently discloses the determined error condition as an alternative limitation that is not currently required by the claim." Office Action, at 5. This point is now moot. Amended claim 1 requires the first machine learning model to "predict an error condition" as the sole analysis step; the previously recited alternative of predicting that "an operation metric will reach a policy-specified CPU threshold within a time period" has been removed. As such, prediction of an error condition is no longer an alternative it is the only recited analysis.
In response, Examiner respectfully disagrees. Examiner notes Applicant’s arguments regarding the machine learning and storing elements are moot because these elements are not considered abstract and therefore were not analyzed under Step 2A-Prong One. However Examiner finds the aforementioned elements do not the take the claim out of the certain methods of organizing human activity and mental processes groupings. Furthermore, Examiner finds analysing the acquired operation metric data to predict an error condition mimics human thought processes of observation, evaluations, judgement and opinion, that can feasibly be performed with pen and paper, where the data interpretation is perceptible in the human mind. Examiner maintains the claim recites an abstract idea.
Regarding the 35 U.S.C. 101 rejection, Applicant submits that the claim as a whole integrates any such exception into a practical application by, at least, improving the functioning of the RPA robot's error-prediction technology. MPEP § 2106.04(d)(1) provides that a claim integrates a judicial exception into a practical application when the specification sets forth an improvement in technology and the claim reflects the disclosed improvement.
The specification describes the technical problem that "[o]rganisations that use RPA software such as software robots (also known as digital workers) have to ensure that the software is performing correctly during runtime" and that "[e]rrors that occur during execution have to be diagnosed and fixed as soon as possible," but that manual maintenance is costly and slow. Specification, at [0003]-[0004]. The specification discloses a solution in which "[t]he first machine learning model may be trained with at least one of: operation metrics of an RPA robot, possible errors in executing an automation workflow by an RPA robot," Specification, at [0031], and in which the method further comprises "storing information comprising at least one of: the acquired data associated with the operation metric of the RPA robot, the predicted or determined error in the execution of the automation workflow" and "training the first machine learning model based on the stored information." Specification, at [0032]. The specification further describes "storing the user input to confirm or reject the determined optimising action, and/or the user input to agree or disagree with the predicted or determined error" and "training the first machine learning model based on the stored user input and/or user feedback." Specification, at [0035].
Amended claim 1 reflects this improvement (see p. 8, Applicant Remarks).
Applicant submits that this is particularly significant in the context of RPA, where automation workflows are executed hundreds or thousands of times. Retraining on data generated by a particular RPA robot and workflow allows the model to learn that robot's and workflow's operational characteristics over time and thereby improve subsequent error prediction. The claim therefore reflects a disclosed improvement to the technology of RPA error prediction, not merely the application of a judicial exception using generic computer components.
In response, Examiner respectfully disagrees. Examiner finds the pending claims are not technological in nature and merely limits the abstract idea to a particular environment and thus fails to add an inventive concept to the claims; see MPEP 2106.05(h). Examiner notes in Ex parte Desjardins, Appeal 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision), the specification identified the improvement to machine learning technology by explaining how the machine learning model is trained to learn new tasks while protecting knowledge about previous tasks to overcome the problem of “catastrophic forgetting”, and the claims reflected the improvement identified in the specification. The improvements identified in the Desjardins specification included disclosure of the effective learning of new tasks in succession in connection with specifically protecting knowledge concerning previously accomplished tasks; allowing the system to reduce use of storage capacity; and the enablement of reduced complexity in the system. Such improvements were tantamount to how the machine learning model itself would function in operation and therefore not subsumed in the identified mathematical calculation. Examiner finds no similar improvement in the present claims. Applicant is describing the use of the additional elements as tools to automate the prediction and remediation of errors validated by a user without any improvement to how the machine learning fundamentally functions, reflecting and/or submitting that the technology used is being improved or there was a technical problem with the technology that the claimed invention solves. Examiner finds the pending claims recite similar limitations to claims the courts have indicated may not be sufficient in showing an improvement in computer-functionality, such as accelerating a process of analyzing audit log data when the increased speed comes solely from the capabilities of a general-purpose computer, FairWarning IP, LLC v. Iatric Sys., 839 F.3d 1089, 1095, 120 USPQ2d 1293, 1296 (Fed. Cir. 2016); Mere automation of manual processes, such as using a generic computer to process an application for financing a purchase, Credit Acceptance Corp. v. Westlake Services, 859 F.3d 1044, 1055, 123 USPQ2d 1100, 1108-09 (Fed. Cir. 2017), A commonplace business method being applied on a general purpose computer, Alice Corp., 573 U.S. at 223, 110 USPQ2d at 1976; Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); Gathering and analyzing information using conventional techniques and displaying the result, TLI Communications, 823 F.3d at 612-13, 118 USPQ2d at 1747-48; see MPEP 2106.05(a)(I) and MPEP 2106.05(a)(II). Examiner finds Applicants arguments aim to improve an existing business process (e.g., improving slow and costly maintenance) and not the functioning of the computer, another technology or computer-related technology. Examiner maintains the claim is directed to an abstract idea.
Regarding the 35 U.S.C. 101 rejection, Applicant states claim 1 recites "causing the RPA robot, while executing the automation workflow, to automatically execute the action command," where the action command comprises concrete remedial actions such as "restarting execution of the automation workflow," "postponing or stopping the execution," "changing an element of the environment in which the RPA robot operates," "changing an operation parameter of the RPA robot," "changing a defining parameter of and/or a stage in the automation workflow," "skipping an optional stage," or requesting different allocation of CPU or memory resources, as recited by claim 1. These are not generic data-processing steps, but they are automatic, in-session actuations of the RPA robot that use the output of the error-prediction model to change the robot's behavior during execution of the workflow.
Applicant also notes a material similarity to USPTO Subject Matter Eligibility Example 47, claim 3 (2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence). In that example, a claim reciting a trained artificial neural network that detects anomalies in network traffic and then takes concrete remedial actions (e.g., detecting a source address, dropping malicious packets, and blocking future traffic) was found eligible at Step 2A, Prong Two because those remedial steps "reflect the improvement described in the background" and "the claim improves the functioning of a computer or technical field."
Similar to Example 47, claim 1 recites a trained model that predicts error conditions and concrete automatic remedial actions taken on the RPA robot during execution. Claim 1 further recites that the model is retrained using confirmed predictions to improve subsequent error prediction. For at least these reasons, Applicant submits that claim 1 as amended integrates any alleged judicial exception into a practical application.
In response Examiner respectfully disagrees. Again, Examiner finds the abovementioned limitations are not technological in nature and merely limits the abstract idea to a particular environment and thus fails to add an inventive concept to the claims; see MPEP 2106.05(h). Examiner finds the action commands are instructions of the abstract idea executable by the generic computer components based on user validation of a predicted error condition. There is no inventive concept added to the claims here.
Furthermore, The Step 2A-Prong Two analysis of Example 47 states (see p. 12, AI-Related SME Examples 47-49 issued in 2024), in part:
According to the background section, existing systems use various detection techniques for detecting potentially malicious network packets and can alert a network administrator to potential problems. The disclosed system detects network intrusions and takes real-time remedial actions, including dropping suspicious packets and blocking traffic from suspicious source addresses. The background section further explains that the disclosed system enhances security by acting in real time to proactively prevent network intrusions.
The claimed invention reflects this improvement in the technical field of network intrusion detection. Steps (d)-(f) provide for improved network security using the information from the detection to enhance security by taking proactive measures to remediate the danger by detecting the source address associated with the potentially malicious packets. Specifically, the claim reflects the improvement in step (d), dropping potentially malicious packets in step (e), and blocking future traffic from the source address in step (f). These steps reflect the improvement described in the background. Thus, the claim as a whole integrates the judicial exception into a
practical application such that the claim is not directed to the judicial exception.
The additional elements in steps (d)-(f), when considered in combination, integrate the abstract idea into a practical application because the claim improves the functioning of a computer or technical field. See MPEP 2106.04(d)(1) and 2106.05(a). The claimed invention reflects this improvement in the technical field of network intrusion detection. Thus, the claim as a whole integrates the judicial exception into a practical application (Step 2A, Prong Two: YES), such that the claim is not directed to the judicial exception. (Step 2A: NO). The claim is eligible.
Examiner finds the pending claims are not similar to the eligible subject matter disclosed in claim 3 of Example 47 because the execution of the action commands is contingent on the user validating the predicted error condition, which is similar to the existing conventional solutions that teach away from the inventive concepts disclosed in Claim 3 of Example 47. Examiner finds the claim is directed to an abstract idea.
Regarding the 35 U.S.C. 101 rejection, Applicant submits that the ordered combination of features in amended claim 1 does not fit the generic "collect-analyze-display" pattern. Claim 1, for example, describes, pre-training the first machine learning model with operation metric data comprising time-stamped event sequences and possible errors, predicting an error condition during execution, receiving user input agreeing with the prediction, automatically executing a remedial action command on the RPA robot, storing the operation metric data, the predicted error condition, and the user input, and training the first machine learning model using that stored data. Such features provide a specific, claim-recited personalization loop that adapts an error- prediction model to a particular RPA robot and workflow. Accordingly, Applicant respectfully submits that this ordered combination amounts to significantly more than any alleged abstract idea.
Independent claim 14 as amended recites a system that performs the method of claim 1 and is eligible for the same reasons. Dependent claims 4, 5, 7, 8, 10, 12, 13, and 15 further narrow the method or system and are eligible at least by virtue of their dependency. Accordingly, Applicant respectfully requests reconsideration and withdrawal of the 35 U.S.C. § 101 rejection.
In response, Examiner respectfully disagrees. Examiner maintains Applicants abovementioned arguments are not technological in nature and merely limits the abstract idea to a particular environment. Specifically, Examiner finds predicting an error condition during execution, receiving user input agreeing with the prediction, automatically executing a remedial action command on the RPA robot are abstract limitations that constitute methods based on certain methods of organizing human activity and mental processes. Storing the operation metric data, the predicted error condition, and the user input, and training the first machine learning model using that stored data are merely rules and instructions implementing the abstract idea using generic computer components and technology. Examiner respectfully reminds Applicant, regardless of the complexity and/or granularity a personalization loop that adapts an error- prediction model to a particular RPA robot and workflow without meaningful limitations within the claims that amount to significantly more than the abstract idea itself is a judicial exception (i.e. abstract idea). Examiner maintains the additional elements recited in the claims do not perform any unconventional functions that can be considered “significantly more” than the judicial exception and amount to no more than a recitation of generic computer elements and technology utilized to perform their generic functions. Applicant has not made any persuasive argument that would alter this analysis. For at least these reasons, the pending claims remain rejected under 35 U.S.C. § 101 as being directed to non-statutory subject matter.
Examiner welcomes Applicant to schedule an interview to discuss the pending 35 U.S.C. 101 rejections.
Applicant’s arguments, see pg. 10-12, filed September 03, 2026, with respect to the rejection(s) of claims 1, 4, 5, 7, 8, 10, and 12-15 under 35 U.S.C. 103 have been fully considered. However, upon further consideration, a new ground(s) of rejection is made. Applicant’s arguments are considered moot because they are directed to newly amended subject matter and do not apply to the combination of references being used in the current rejection. Please refer to the 35 U.S.C. 103 rejection for further explanation and rationale.
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, 3-8 and 10-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: The claimed subject matter falls within the four statutory categories of patentable subject matter.
Claims 1, 3-8, and 10-12 are directed towards a method, claims 13 is directed towards a computer-readable storage medium and claims 14 and 15 are directed towards a system, which are among the statutory categories of invention.
Step 2A – Prong One: The claims recite an abstract idea.
The claims recite validating error conditions and determining actions commands for a Robotics Process Automation robot based on user input, policy and operation metrics.
Claim 1 recites limitations directed to an abstract idea based on certain methods of organizing human activity and mental processes. Specifically, analysing the acquired operation metric data to predict an error condition, determining, based on a policy and the predicted error condition, an action command and receiving user input to agree with the predicted error condition which is an automation of a business process constituting methods based on commercial interactions, as well as methods based on observations, evaluations, judgements and/or opinion that can be performed mentally by a combination of the human mind and a human using pen and paper. The recitation of the RPA robot, operating system and processors and first machine learning model does not take the claim out of the certain methods of organizing human activity and mental processes groupings. Thus the claim recites an abstract idea. Claims 13 and 14 recite certain method of organizing human activity and mental processes for similar reasons as claim 1.
Step 2A – Prong Two: The judicial exception is not integrated into a practical application.
The judicial exception is not integrated into a practical application. In particular, claim 1 recites acquiring, during execution of an automation workflow by the RPA robot, operation metric data via one or more sensors, the operation metric data comprising a time-stamped sequence of events in the automation workflow and storing the operation metric data, the predicted error condition, and the user input, which are limitations considered an insignificant extra-solution activity of collecting and delivering data; see MPEP 2106.05(g). Claim 1 further recites an RPA robot, operating system and processors at a high-level of generality such that they amount to no more than generic computer components used as tools to apply the action command execution instructions of the abstract idea; see MPEP 2106.05(f). Additionally, wherein the action command comprises at least one of: restarting execution of the automation workflow by the RPA robot, postponing or stopping the execution of the automation workflow by the RPA robot, changing an element of the environment in which the RPA robot operates, changing an operation parameter of the RPA robot, changing a defining parameter of and/or a stage in the automation workflow, skipping an optional stage in the automation workflow during execution, requesting different allocation of central processing unit, CPU, resources, requesting different allocation of memory resources, and requesting extra workload to be performed by the RPA robot are not technological in nature and merely limits the abstract idea to a particular technological environment; see MPEP 2106.05(h). Claim 1 further recites wherein the first machine learning model is pre-trained with operation metric data comprising time-stamped sequences of events in automation workflows of an RPA robot and possible errors in executing automated workflows by an RPA robot and training the first machine learning model using the stored operation metric data, the stored predicted error condition, and the stored user input. The general use of a machine learning technique does not provide a meaningful limitation to transform the abstract idea into a practical application. Therefore, the machine learning model disclosed in the claim is solely used as a tool to perform the instructions of the abstract idea. Claim 1 as a whole, looking at the additional element individually and in combination with the remaining claim limitations, does not integrate the judicial exception into a practical application and therefore is directed to an abstract idea. The computer-readable storage medium comprising executable instructions by a computer recited in claim 13 and the system comprising one or more processors and a memory storing instructions executable by the one or more processor in claim 14 also amount to no more than mere instructions to apply the exception using generic computer components; see MPEP 2106.05(f). Thus, the additional elements recited in claims 13 and 14 do not integrate the abstract idea into practical application for similar reasons as claim 1.
Step 2B: The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements in the claims other than the abstract idea per se, including system comprising one or more processors and a memory storing instructions executable by the one or more processor, operating system, a non-transitory computer-readable storage medium and RPA robot, amount to no more than a recitation of generic computer elements utilized to perform generic computer functions, such as receiving or transmitting data over a network, e.g., using the Internet to gather data, buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); electronic recordkeeping, Ultramercial, 772 F.3d at 716, 112 USPQ2d at 1755 (updating an activity log) and storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93; see MPEP 2106.05(d)(II). The machine learning model recited in the claims is disclosed at a high-level of generality (see at least Specification [0037]) and does not amount to significantly more than the abstract idea. Viewed as a whole, these additional claim elements do not provide meaningful limitations to transform the abstract idea into a patent eligible application of the abstract idea such that the claims amount to significantly more than the abstract idea itself. Therefore, since there are no limitations in the claim that transform the abstract idea into a patent eligible application such that the claim amounts to significantly more than the abstract idea itself, the claims are rejected under 35 U.S.C. § 101 as being directed to non-statutory subject matter.
§ 101 Analysis of the dependent claims.
Regarding the dependent claims, dependent claim 15 recite wherein the system is providable as an external component to the automation workflow, which is not technological in nature and merely limits the abstract idea to a particular environment. Additionally, claims 4, 5, 7, 8 and 10 and 12 recite steps that further narrow the abstract idea. No additional elements are disclosed in the dependent claims that were not considered in the independent claims. Therefore claims 4, 5, 7, 8, 10, 12 and 15 do not provide meaningful limitations to transform the abstract idea into a patent eligible application of the abstract idea such that the claims amount to significantly more than the abstract idea itself.
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 (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 4, 5, 7, 8, 10 and 13-15 are rejected under 35 U.S.C. 103 as being unpatentable over Ghag et al., U.S. Publication No. 2023/0039566 [hereinafter Ghag], and further in view of Siebel et al., U.S. Publication No. 2023/0291755 [hereinafter Siebel].
Referring to Claim 1, Ghag teaches:
A computer-implemented method for training a machine learning model for predicting error conditions of a Robotics Process Automation, RPA, robot, the method comprising:
acquiring, during execution of an automation workflow by the RPA robot, operation metric data via one or more sensors, the operation metric data comprising a time-stamped sequence of events in the automation workflow (Ghag, [0058]), “monitor the RPA components and all the dependent components using the monitor module 254, detect anomaly using the metric processor 256, and remediate the detected anomaly using the remediation action module 258. The monitoring may either be a scheduled monitoring or a triggered monitoring, wherein the monitoring may monitor/scan resources based on observable metrics…”; (Ghag, [0059]), “the monitor module 254 may fetch the observable metrics configured for each resource from the RPA metadata configuration database 222… The script execution engine 260 may fetch the script from the script repository 262 using the Script ID and execute the scripts centrally, by sending one or more instructions to RPA platform 202 to obtain values for the observable metrics (for example, CPU utilization, response time etc.). Once the script is executed, the script execution response parameters are returned to the script execution engine 260 and then to the monitor module 254 to prepare the metric message…”; (Ghag, [0049]-[0050]);
analysing, by one or more processors and a first machine learning model, the acquired operation metric data to predict an error condition, wherein the first machine learning model is pre-trained with operation metric data comprising time-stamped sequences of events in automation workflows of an RPA robot and possible errors in executing automated workflows by an RPA robot (Ghag, [0037]), “detecting at least one anomaly by validating the values of the one or more observation metrics. The determination of the values of the one or more observation metrics comprises, querying the one or more observation metrics of the one or more resources and at least one script associated with each of the one or more metrics; executing the at least one script to fetch the values of the one or more observation metrics from the one or more resources; and generating a metric message comprising the values for each of the one or more observation metrics. The detection of the at least one anomaly, comprising, parsing the metric message to obtain values of the one or more observation metrics; comparing the values of the one or more observation metrics against a threshold value for each of the one or more observation metrics; and determining the values of the one or more observation metrics as anomaly if the values of the one or more observation metrics breach the threshold value”; (Ghag, [0052]), “model training engine”; (Ghag, [0054]), “…A model can be trained for a particular resource observable metric on different variables such as time of day, number of records to be processed by the bot etc. to predict the threshold value which could be in terms of bot queue length, bot processing time etc. Also, the model building engine 238 may receive user input from model training interface 232 through observable metric threshold configuration module 234 to tune the dynamic threshold parameters”; (Ghag, [0073]), “training machine learning models may comprise, receiving metrics data from metrics data store associated with the automation engine, attributes of the one or more resources from the RPA metadata configuration database and historic unstructured log data from log data store associated with the RPA platform… extracted error patterns may be provided as input to the to train the machine learning models”; (Ghag, [0061]), “metric processor 256 may execute anomaly detection on the metric values. Anomaly may be detected by validating the metric values against a threshold value…dynamic threshold, the system may be configured to automatically determine upper threshold and/or lower threshold based on the analysis of historic metrics data and historic log data to generate and train machine learning models”; (Ghag, [0038]; [0060]; [0072]; [0081]);
determining, based on a policy and the predicted error condition, an action command (Ghag, [0038]), “detected at least one anomaly is remediated by identifying at least one automated remediation action comprising sequence of instructions and executing the identified at least one automated remediation action causing the remediation of the detected at least one anomaly”; (Ghag, [0062]), “the remediation action module 258 may be configured to execute Remediation Action Plan for the anomaly detected by the metric processor 256. Based on the observation through the observable metrics for which an anomaly has been detected, the remediation action module 258 may determine appropriate Remediation Action Plan to be executed to resolve the anomaly…”;
causing the RPA robot, while executing the automation workflow, to automatically execute the action command (Ghag, [0062]), “… A Remediation Action Plan may comprise series of action steps which will be executed in sequence by the script execution engine 260 as configured. Each action is either linked with a script or a workflow in case of executing more complex orchestration logic. When the anomaly is detected by the metric processor 256, the remediation action module 258 may identify the remediation action plan to be executed by sending one or more instructions to the script execution engine 260. The Remediation Action Plan will start executing the action steps wherein each action steps contain details of the script/workflow to be executed”; (Ghag, [0063]), “remediation action module 258 may send one or more instructions to the script execution engine 260 along with the script ID of the identified Remediation Action Plan, the observable metrics and the values, for example, component name, IP address and/or user credentials, so that script execution engine 260 execute the action steps centrally. The script execution engine 260 may fetch the script to be executed from the script repository 262 based on the script ID and then execute the actions steps which are part of fetched script which causes the script execution engine 260 to send one or more instructions to the RPA platform 202, which causes the resolution of anomaly detected in the RPA platform…”
storing the operation metric data, the predicted error condition, and the user input (Ghag, [0063]), “the resolution (also termed as remediation) may indicate change in state of operation of resources—either to bring back the resource to a healthy state i.e., a state which will not violate/breach the defined threshold, or to start/restart the resource if the resource is shut down/non-functional. After the completion of all the action steps present in the script, a status is updated for the anomaly detected in the operations database 266 and a notification may be sent to the impacted parties/user. If the execution of remediation action plan fails in any of the action steps then the error message may be logged, status may be updated in the operations database 266”; (Ghag, [0082]); and
wherein the action command comprises at least one of: restarting execution of the automation workflow by the RPA robot, postponing or stopping the execution of the automation workflow by the RPA robot, changing an element of the environment in which the RPA robot operates, changing an operation parameter of the RPA robot, changing a defining parameter of and/or a stage in the automation workflow, skipping an optional stage in the automation workflow during execution, requesting different allocation of central processing unit, CPU, resources, requesting different allocation of memory resources, and requesting extra workload to be performed by the RPA robot (Ghag, [0063]), “the resolution (also termed as remediation) may indicate change in state of operation of resources—either to bring back the resource to a healthy state i.e., a state which will not violate/breach the defined threshold, or to start/restart the resource if the resource is shut down/non-functional”; (Ghag, [0071]; [0051]).
Ghag teaches an error type labeling module receiving inputs from a user which allows the user to mark/identify patterns identified by the error pattern extraction engine as errors (see par. 0056), an analytics engine may be configured to notify user through a notification displayed at the computing device associated with the user when an anomaly is detected (see par. 0065) and after the completion of all the action steps present in the script, a status is updated for the anomaly detected in the operations database (see par. 0063), but Ghag does not explicitly teach:
receiving user input to agree with the predicted error condition; and
training the first machine learning model using the stored operation metric data, the stored predicted error condition, and the stored user input.
However Siebel does not explicitly teach:
receiving user input to agree with the predicted error condition (Siebel, [0062]-[0063]), “…the labels that are used by the supervised anomaly detection operation 504 may be provided by the enterprise running the AI platform, generated using results from the unsupervised anomaly detection operation 502 (which may be referred to as bootstrapping), generated using internal “red-teaming” efforts where trusted parties attack the system 100 and record the time and nature of their attacks, or obtained from users of the AI platform (such as through feedback from security analysts who analyze detected anomalies)… feedback from one or more security analysts or other users who accept and reject alerts for detected anomalies”; (Siebel, [0088]; [0051]); and
training the first machine learning model using the stored operation metric data, the stored predicted error condition, and the stored user input (Siebel, [0063]), “additional training data may also take the form of feedback from one or more security analysts or other users who accept and reject alerts for detected anomalies, where the feedback can be used during retraining or reinforcement learning”; (Siebel, [0112]), “Retraining and/or reinforcement learning may optionally be used to update the unsupervised and supervised anomaly detection components at step 1308. This may include, for example, the processing device 202 of the application server 106 using additional training data (such as feedback from human personnel related to prior anomaly detection results) to update the machine learning models used for unsupervised and supervised anomaly detection. This step can occur at any suitable time(s), such as periodically, irregularly, on demand, or in any other suitable manner”; (Siebel, [0082]; [0027]).
At the time the invention was filed, it would have been obvious to a person of ordinary skill in the art to have modified user input and operations database in Ghag to include the agreement and model training limitations as taught by Siebel. The motivation for doing this would have been to improve the method of automated detection and automated remediation of anomalies in a Robotic Process Automation (RPA) environment in Ghag (see par. 0009) to include the results of supporting continuous improvement of the machine learning models' performances over time (see Siebel par. 0063).
Referring to Claim 4, Ghag in view of Siebel teaches the method according to claim 1. Ghag further teaches:
wherein the predicted error condition in the execution of the automation workflow by the RPA robot corresponds to an abnormal behaviour in the execution of the automation workflow (Ghag, [0034]), “the observables may be monitored for anomalies based on Anomaly Detection techniques, wherein one of the techniques is using thresholds (or thresholding models or threshold models). In the thresholding models, threshold values for the metrics are used by the system to identify anomalies in the behavior of the metrics being monitored… For example, if CPU utilization value is greater than 75%, it may be considered as breach if the threshold is set as 75%, which is a rule-based threshold. In case of dynamic threshold, the value of the metric may be derived from other operating parameters in the environment. For example, say during a certain period of the last 3 months historical data, the system has learned that CPU utilization on Thursdays at 6 PM is usually 40%, whereas CPU utilization on Thursdays at 10 AM is 70%. These CPU utilization values are learned by a machine learning system based on parameters which in this case being ‘day of the week’ and ‘time’. So, the value of CPU utilization may change in the future based on operations and system will automatically learn about this change in CPU utilization by understanding the day of the week/time metric of the past historical data”; (Ghag, [0058]), “the automation engine 208 may be configured to monitor the RPA components and all the dependent components using the monitor module 254, detect anomaly using the metric processor 256, and remediate the detected anomaly using the remediation action module 258. The monitoring may either be a scheduled monitoring or a triggered monitoring, wherein the monitoring may monitor/scan resources based on observable metrics such as, but not limited to bot runtime, bot queue length, virtual machine CPU utilization, Control Tower service status etc.”; (Ghag, [0038]; [0064]).
Referring to Claim 5, Ghag in view of Siebel teaches the method according to claim 1. Ghag further teaches:
wherein the action command further comprises at least one of: a message corresponding to the predicted error condition and an action recommendation corresponding to the predicted error condition (Ghag, [0061]), “If the validation determines that if there is a breach of threshold, then the captured observable metric may be marked as an anomaly and an event/alert may be raised to take further actions which could be sending out a notification to users and/or triggering an automated remediation process as configured”; (Ghag, [0065]), “analytics engine 210 may be configured to notify user through a notification displayed at the computing device associated with the user when an anomaly is detected… the analytics engine 210 may be configured to display the status of the remediation action performed. A notification module of the analytics engine 210 may recognize the type of notification to be sent, load the relevant notification template from a list of template documents stored in a database and fill the template with specific data of the relevant resource/observations/actions for the user”.
Referring to Claim 7, Ghag in view of Siebel teaches the method according to claim 1. Ghag further teaches:
wherein analysing the acquired operation metric data further comprising using a deterministic algorithm (Ghag, [0050]), “the observables may be monitored for anomalies based on Anomaly Detection techniques, wherein one of the techniques is using thresholds (or thresholding models or threshold models). In the thresholding models, threshold values for the metrics are used by the system to identify anomalies in the behavior of the metrics being monitored. These threshold values may either be static (rule-based, also referred as ‘deterministic’) i.e. same value in any operating condition with an assumption that no underlying environment variability will have an impact on the resource behavior; or dynamic (or adaptive, also referred as ‘non-deterministic’) where threshold values are computed dynamically for parts of the larger dataset which have dependency on underlying dependent variables such as time of the day, transaction volume, CPU utilization, etc. which may impact the behavior of the resources”; (Ghag, [0069]; [0081]).
Referring to Claim 8, Ghag in view of Siebel teaches the method according to claim 1. Ghag further teaches:
further comprising receiving at least one of: a user input to enable or disable automatic execution of the determined action command, and a user input indicating user feedback subsequent to execution of the action command (Ghag, [0079]), “User can activate/de-activate any specific Resource (412); (b) User can add resource, click on plus icon (418) and a popup will be displayed with three options—(i) Add Parameter; (ii) Add Resource; and (iii) Add Observable-Remediation plan; (c) User can Select/Deselect (420) the Observable-Remediation plan”; (Ghag, [0048]), “One or more monitoring and remediation action plans may be set up by the user as a part of initial configuration by the user. The ‘activate’ process may comprise steps such as configuring credentials for script execution and other parameters, activating the monitoring task and activating remediation plans for the RPA components, and review the RPA resources which are up and functioning”; (Ghag, [0059]).
Referring to Claim 10, Ghag in view of Siebel teaches the method according to claim 1. Ghag further teaches:
wherein the policy is a rule-based policy that dictates a corresponding condition for comparison with the operation metric and an action to be carried out based on results of the comparison between the condition and the operation metric (Ghag, [0050]), “the observables may be monitored for anomalies based on Anomaly Detection techniques, wherein one of the techniques is using thresholds (or thresholding models or threshold models). In the thresholding models, threshold values for the metrics are used by the system to identify anomalies in the behavior of the metrics being monitored. These threshold values may either be static (rule-based, also referred as ‘deterministic’) i.e. same value in any operating condition with an assumption that no underlying environment variability will have an impact on the resource behavior… For example, if CPU utilization value is greater than 75%, it may be considered as breach if the threshold is set as 75%, which is a rule-based threshold”; (Ghag, [0073]), “One or more error patterns from the structured format data may be extracted using regex rules for error pattern detection. The regex rules may be either static regex rules that are pre-defined by the user or dynamic regex rules wherein the system is able to automatically extract error patterns, for example, using longest common sequence algorithm”; (Ghag, [0038]; [0081]).
Referring to Claim 13, Ghag teaches:
A non-transitory computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the method of claim 1 (Ghag, [0080]; [0082]).
Referring to Claim 14, Ghag teaches:
A system for training a machine learning model for predicting error conditions of a Robotics Process Automation, RPA robot, the system comprising one or more processors and a memory storing instructions, which when executed by the one or more processors, cause the system to perform the method of claim 1 (Ghag, [0041]-[0042]).
Referring to Claim 15, Ghag in view of Siebel teaches the system according to claim 14. Ghag further teaches:
wherein the system is providable as an external component to the automation workflow (Ghag, [0037]), “One or more RPA platforms and computing devices communicatively coupled to the one or more RPA platforms together forms an RPA environment. The method comprising, discovering one or more resources in an RPA platform. The discovered one or more resources on the RPA platform may be monitored”; (Ghag, [0056]), “The generated and trained models may be stored in the models storage device 252. As, the models storage device 252 may be communicatively coupled to the configuration interface 214, it may allow user to associate an error type with remediation action at the configuration interface 214. After the configurations of all RPA components and its dependencies, and creation of models based on historical data available from the RPA systems, the automation engine 208 may monitor and detect anomalies”; (Ghag, [0039]).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Ghag et al., U.S. Publication No. 2023/0039566 [hereinafter Ghag], in view of Siebel et al., U.S. Publication No. 2023/0291755 [hereinafter Siebel], and further in view of Rajagopalan et al., U.S. Publication No. 2022/0121982 [hereinafter Rajagopalan].
Referring to Claim 12, Ghag in view of Siebel teaches the method according to claim 1. Ghag teaches Once the models are built and trained, they are published and exposed as API which can be used by the automation engine at runtime (see par. 0056), but Ghag does not explicitly teach:
wherein acquiring the operation metric of the RPA robot during execution of an automation workflow comprises acquiring first data associated with a first session at a runtime resource corresponding to the automation workflow, and wherein the method further comprises: acquiring second data associated with an operation metric of the RPA robot during execution of a second automation workflow, wherein the second data are associated with a second session at the runtime resource; and determining a second command based on a second policy and the acquired second data.
However Rajagopalan teaches:
wherein acquiring the operation metric of the RPA robot during execution of an automation workflow comprises acquiring first data associated with a first session at a runtime resource corresponding to the automation workflow (Rajagopalan, [0077]-[0079]), and wherein the method further comprises:
acquiring second data associated with an operation metric of the RPA robot during execution of a second automation workflow, wherein the second data are associated with a second session at the runtime resource (Rajagopalan, [0079]), “Ongoing monitoring and evaluation of the workload executions and/or the computing system environment may comprise monitoring, logging, and/or analyzing real-time/runtime configuration parameter settings, values, and/or real-time/runtime performance metrics”; (Rajagopalan, [0077]-[0078]); and
determining a second command based on a second policy and the acquired second data (Rajagopalan, [0079]), “Logged real-time/runtime data may be validated against defined policies and/or templates. Deviations from acceptable values and ranges may be flagged to notify an administrative function of the deviations, initiate an analysis, and/or initiate a determination of one or more corrective actions that may be taken to bring the performance metrics back into conformance with the policies and/or templates”; (Rajagopalan, [0078]; [0080]), Examiner notes “Mere duplication of parts has no patentable significance unless a new and unexpected result is produced”; see MPEP 2144.04.
At the time the invention was filed, it would have been obvious to a person of ordinary skill in the art to have modified the runtime in Ghag to include the workflow and action limitations as taught by Rajagopalan. The motivation for doing this would have been to improve the method of automated detection and automated remediation of anomalies in a Robotic Process Automation (RPA) environment in Ghag (see par. 0009) to include the results of optimizing resource utilization, bandwidth utilization, and efficient operations of the enterprise computing infrastructure (see Rajagopalan par. 0002).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
NG et al. (US 20220292661 A1) – An automated inspection system includes a machine-vision inspection device, a repair station and an artificial intelligence module connected to the machine-vision inspection device, and a processing platform connected to the repair station and the artificial intelligence module. The machine-vision inspection device obtains an image of a to-be-inspected object and displays the image on an operation screen of the repair station, the artificial intelligence module retrieves the image of the to-be-inspected object to perform inference for classification, the processing platform automatically overlaps result data of classification inference for the image of the to-be-inspected object on the operation screen of the repair station, and intercepts and transmits result data of manual review, which is inputted by an operator through a user interface, to the repair station. The processing platform then transmits the result data of manual review to the artificial intelligence module as the data for automatically re-training a classification model.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Crystol Stewart whose telephone number is (571)272-1691. The examiner can normally be reached 9:00am-5:00pm.
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, Patty Munson can be reached on (571)270-5396. 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.
/CRYSTOL STEWART/Primary Examiner, Art Unit 3624
