Engine Monitoring Model for Early Failure Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing engine monitoring systems are inaccurate in predicting damage and optimizing performance due to unforeseen operations, often detecting damage at catastrophic levels or reducing effectiveness when faced with unforeseen actions, and fail to provide timely maintenance recommendations.

Innovation Solution

A remote monitoring system that receives historical usage data to train an engine monitoring model, identifies operational ranges for operating parameters, predicts engine failure, and adjusts configurations to prevent failures by analyzing usage patterns and sending notifications for maintenance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the engine is operated aggressively for higher performance, then performance is improved, but durability deteriorates due to faster damage accumulation

Engineering Contradiction:

Improveengine performanceVSAvoidengine durability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary damage assessment by continuously monitoring engine parameters and comparing them against a damage model before catastrophic failure occurs. This allows operators to take preventive actions (reduce load, schedule maintenance) before the damage becomes critical, resolving the contradiction by enabling informed trade-offs between performance and durability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system establishes a feedback loop where engine operating parameters are continuously measured, fed into a damage model, and the resulting damage assessment is returned to operators. This feedback mechanism enables real-time adjustments to operating conditions, allowing optimization of both performance and durability by preventing operation in damaging regimes.

Inventive Principle:

Principle #23Feedback

2Measurement precision

If a monitoring model is used to detect engine damage, then damage detection capability is improved, but accuracy deteriorates when unforeseen operations cause the model to be inaccurate

Engineering Contradiction:

Improvedamage detection accuracyVSAvoidmodel adaptability to unforeseen operations

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The damage model is designed to be dynamic rather than static. It continuously adapts to new operating conditions by incorporating real-time sensor data and updating damage assessments. This dynamic approach allows the model to remain accurate even when encountering unforeseen operations, resolving the contradiction between detection precision and adaptability.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system monitors multiple engine parameters simultaneously (temperature, pressure, vibration, etc.) and uses changes in these parameters over time to assess damage. By tracking parameter trends rather than relying on fixed thresholds, the model maintains accuracy across diverse operating conditions and adapts to unforeseen operations.

Inventive Principle:

Principle #35Parameter changes

3Object-affected harmful factors

If the monitoring model detects damage at catastrophic levels, then damage severity is improved for safety, but useful life deteriorates due to late detection

Engineering Contradiction:

Improvedamage severity detectionVSAvoidengine useful life

Core Design Contradiction:

Object-affected harmful factorsVSDuration of action of stationary object

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system rushes through the damage detection process by continuously monitoring multiple parameters in real-time, skipping the wait time associated with traditional periodic inspection methods. This continuous monitoring approach detects damage at early stages rather than at catastrophic levels, extending engine useful life while maintaining safety.

Inventive Principle:

Principle #21Skipping (Rushing through)

Data Source

PatentUS11650580B2Monitoring system for engine performance and failure prediction

Publication Date: 2023.05.16 CATERPILLAR INC

AI summary

In some implementations, a remote monitoring system may receive historical usage data associated with a plurality of engines that are associated with a plurality of respective machines. The remote monitoring system may train an engine monitoring model to identify a usage profile that indicates potential failure by identifying an operational range of an operating parameter according to an operating profile. The remote monitoring system may receive, from a machine, usage data that includes a measurement of the operating parameter for an engine of the machine, and determine that the engine is configured to operate according to the operating profile. The remote monitoring system may predict, based on determining that the engine is configured to operate according to the operating profile and based on the measurement and the operational range, that the engine is likely to fail within a certain time period.