Autonomous System Anomaly Remediation via ML Inference

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing monitoring and logging tools in IT operations fail to capture information related to remediation actions, leading to inconsistent and time-consuming human-driven processes that negatively impact platform availability and increase support costs.

Innovation Solution

The implementation of cognitively assorted machine learning algorithms that identify system anomalies by generating inference models and using machine reinforcement learning to determine remediation actions, predicting adverse states, and automatically executing actions to address these issues.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If human-driven remediation is used to identify and address system issues, then flexibility in handling diverse problems is improved, but consistency and speed of remediation deteriorate

Engineering Contradiction:

Improveflexibility in handling diverse problemsVSAvoidconsistency of remediation

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system uses machine learning models to autonomously identify system anomalies and select appropriate remediation actions without human intervention. The inference model automatically analyzes system log data, identifies adverse states and root causes, while the action policy automatically selects and executes remediation actions, enabling the system to serve itself and achieve consistent, scalable remediation across diverse problems

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system implements a feedback loop where the inference model continuously monitors system log data, predicts adverse states, identifies root causes, and the action policy executes remediation actions based on this analysis. The system learns from historical data and feedback to improve its anomaly detection and remediation capabilities over time, maintaining both consistency and adaptability

Inventive Principle:

Principle #23Feedback

2Adaptability or versatility

If human-driven remediation is used to address system issues, then complex problem-solving capability is improved, but time consumption and support costs increase

Engineering Contradiction:

Improvecomplex problem-solving capabilityVSAvoidtime consumption for remediation

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system pre-trains inference models using historical system log data before deployment. The models are prepared in advance to quickly identify anomalies and predict adverse states when deployed, eliminating the need for time-consuming human analysis during actual remediation events while maintaining complex problem-solving capability

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system replaces human engineers with machine learning-based automated remediation. The inference model and action policy constitute an automated mechanical system that processes system log data and executes remediation actions without human intervention, dramatically reducing time consumption and support costs while maintaining or improving remediation quality

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If automated remediation systems are implemented, then speed and consistency of remediation are improved, but system complexity and development costs increase

Engineering Contradiction:

Improveconsistency of remediationVSAvoidsystem complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The automated remediation system is segmented into distinct functional modules: the inference model for anomaly detection and root cause identification, and the action policy for remediation selection and execution. This modular architecture separates concerns, making the system easier to develop, maintain, and update while maintaining high consistency in remediation operations

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system introduces machine learning models as intermediaries between system log data and remediation actions. The inference model acts as an intermediary to translate raw log data into meaningful anomaly detections and root cause identifications, while the action policy serves as an intermediary to translate system states into appropriate remediation actions, simplifying the overall system architecture

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11514347B2Identifying and remediating system anomalies through machine learning algorithms

Publication Date: 2022.11.29 DELL PROD LP

AI summary

Methods, apparatus, and processor-readable storage media for identifying and remediating anomalies through cognitively assorted machine learning algorithms are provided herein. A computer-implemented method includes: identifying, using system log data, a target variable based at least in part on correlations between a set of performance indicators of a system and the target variable, and threshold values for the performance indicators relative to the target variable; generating an inference model to predict when the system will enter an adverse state and identify one or more root causes of the system entering the adverse state; using machine reinforcement learning to determine an action policy including actions that remediate the adverse state; predicting that the system will enter the adverse state by applying the inference model to further system log data; and automatically executing one or more actions of the action policy in response to the prediction.