Industrial Machine Failure Prediction With Corrective Recommendations

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing monitoring systems for industrial machines fail to predict failures accurately, often resulting in unnecessary maintenance, downtime, and lost revenue due to reliance on predetermined rules and periodic testing, which ignore vast amounts of collected data and require specialized operators.

Innovation Solution

A system that monitors industrial machine behavioral models to identify characteristics associated with previous failures, determines corrective solutions, and generates notifications when similar characteristics are detected, using a processing circuitry and memory to analyze sensory data and predict forthcoming failures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If periodic testing at predetermined intervals is used, then maintenance can be scheduled, but failures may not be prevented and premature replacement occurs

Engineering Contradiction:

Improvemachine reliabilityVSAvoiddowntime

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary action by continuously monitoring machine data and predicting failures before they occur. The machine learning model analyzes current data patterns against historical failure data to identify early warning signs, enabling maintenance to be scheduled at the optimal time - neither too early (premature replacement) nor too late (failure occurs).

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system implements feedback by continuously comparing real-time machine data with historical failure patterns. The machine learning model learns from each prediction and outcome, refining its accuracy over time. This closed-loop feedback mechanism enables the system to adapt to changing machine conditions and improve prediction precision.

Inventive Principle:

Principle #23Feedback

2Device complexity

If predetermined rules are used for monitoring, then simple implementation is achieved, but vast amounts of collected data are ignored

Engineering Contradiction:

Improvemonitoring system complexityVSAvoidunused data

Core Design Contradiction:

Device complexityVSLoss of information

Solution Approach 1:

The system replaces traditional mechanical rule-based monitoring with an intelligent machine learning-based system. Instead of using predefined thresholds and rules that ignore nuanced patterns, the machine learning model automatically learns complex relationships in the data, utilizing all collected information to make predictions.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If specialized operators are required, then accurate analysis can be performed, but human error and cost increase

Engineering Contradiction:

Improvefailure prediction accuracyVSAvoidoperator requirement

Core Design Contradiction:

Measurement precisionVSEase of operation

Solution Approach 1:

The system implements self-service by automatically performing failure prediction and maintenance scheduling without requiring specialized human operators. The machine learning model continuously analyzes machine data, identifies failure risks, and generates maintenance recommendations autonomously, eliminating human error and reducing operational complexity.

Inventive Principle:

Principle #25Self-service

4Measurement precision

If all collected data is analyzed, then accurate failure prediction is achieved, but computing resources are consumed

Engineering Contradiction:

Improvefailure prediction accuracyVSAvoidcomputing resource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system applies partial action by focusing computational resources on analyzing only the most relevant data features for failure prediction. The machine learning model identifies and prioritizes key indicators from the vast dataset, performing deep analysis only on critical parameters while using lighter weighting for less relevant data, thus balancing accuracy with resource efficiency.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS11822323B2Providing corrective solution recommendations for an industrial machine failure

Publication Date: 2023.11.21 AB SKF SKF PATENT DEPARTMENT

AI summary

A system and method for providing a corrective solution recommendation for an industrial machine failure, the method including: monitoring a plurality of segments of at least an industrial machine behavioral model to identify a first segment having at least a first set of characteristics associated with a previous machine failure; determining a corrective solution recommendation that solved the previous machine failure; identifying at least a second set of characteristics associated with a second segment; and generating a notification comprising the corrective solution recommendation when the second set of characteristics is determined to be similar to the first set of characteristics above a predetermined threshold.