Hardware Failure Prediction via Segmented ML Windows

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for determining hardware device failure are often inaccurate, time-consuming, and resource-intensive, requiring multiple models to identify specific component failures, which limits their effectiveness.

Innovation Solution

An automated method using a machine learning hardware device that retrieves operational data, determines differing time frame observation windows, analyzes the data, and generates prediction software applications to forecast hardware or software malfunctions and specific component failures, improving failure prediction accuracy and efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple models are used to determine specified component failure, then measurement precision is improved, but device complexity increases

Engineering Contradiction:

Improvefailure prediction accuracyVSAvoidprocess complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the failure prediction process into two distinct software applications: an apparatus malfunction prediction software application for predicting hardware or software malfunctions, and a component prediction software application for predicting specified component failures. Each application is trained on different feature sets optimized for its specific prediction task, thereby achieving high measurement precision while avoiding the complexity of a single monolithic multi-model system.

Inventive Principle:

Principle #1Segmentation

2Measurement precision

If multiple models are used to determine specified component failure, then measurement precision is improved, but loss of time increases

Engineering Contradiction:

Improvefailure prediction accuracyVSAvoidprediction time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

By segmenting the prediction task into two specialized software applications with dedicated feature sets, the system can execute each application efficiently for its specific purpose. The apparatus malfunction prediction software application quickly identifies potential failures, and the component prediction software application then focuses on identifying specific failed components, reducing overall prediction time compared to running multiple general-purpose models.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The apparatus malfunction prediction software application performs preliminary identification of potential hardware or software malfunctions before the component prediction software application attempts to identify specific failed components. This preliminary action filters the analysis scope, allowing the second application to focus computational resources on relevant components only, thereby reducing total prediction time while maintaining high precision.

Inventive Principle:

Principle #10Preliminary action

3Measurement precision

If multiple models are used to determine specified component failure, then measurement precision is improved, but use of energy increases

Engineering Contradiction:

Improvefailure prediction accuracyVSAvoidcomputational resource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent divides the computational workload into two specialized software applications, each trained on specific feature sets relevant to its prediction task. This segmentation allows each application to be more computationally efficient than a single model attempting to handle all prediction aspects, thereby reducing overall energy consumption while maintaining or improving measurement precision.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system applies partial action by using different feature sets for different prediction tasks rather than computing all possible features for all predictions. The apparatus malfunction prediction software application uses features relevant to hardware/software malfunction detection, while the component prediction software application uses features specifically relevant to component failure identification, reducing unnecessary computational energy expenditure.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11301315B2Automated hardware failure prediction framework

Publication Date: 2022.04.12 KYNDRYL INC

AI summary

A method and system for improving an automated hardware apparatus failure prediction system is provided. The method includes automatically retrieving operational data associated with operation of a hardware device being monitored for potential failure. Differing time frame software windows associated with observing operational data and hardware device are determined and the operational data is analyzed. In response, an apparatus malfunction prediction software application and a component prediction software application is generated. Features associated with execution of the software applications are generated and a first group of features are added to software code of the apparatus malfunction prediction software application. A second group of features are additionally added to software code of the component prediction software application.