Oil Debris Monitoring With Adaptive Learning for False Alarm Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current oil debris monitoring systems in gas turbine engines face challenges in accuracy due to flow volatility and changes in engine components over time, leading to inconsistent and unreliable particle detection.

Innovation Solution

Implementing a method and system for adaptive learning that uses oil debris monitoring sensor data and fleet data to detect anomalies, applying a supervised machine learning classification algorithm like a support vector machine to improve real-time particle detection accuracy, and continuously fine-tuning the algorithm with field maintenance feedback to adjust detection parameters and maintenance actions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional oil debris monitoring is used, then the system is simple to operate, but the measurement precision deteriorates due to flow volatility and component changes

Engineering Contradiction:
Improveparticle detection accuracyVSAvoidmonitoring system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The monitoring system transitions from static threshold-based detection to dynamic adaptive learning that automatically adjusts to changing oil flow conditions and component wear patterns over time, improving measurement precision without requiring manual recalibration

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system incorporates feedback loops where detection results are continuously compared against learned patterns, and the algorithm adapts its parameters based on accumulated data, enabling the system to maintain high accuracy despite variations in operating conditions

Inventive Principle:
Principle #23Feedback

2Reliability

If fixed detection thresholds are used, then the device complexity is low, but the reliability deteriorates due to false alarms from flow volatility

Engineering Contradiction:
Improvedetection reliabilityVSAvoidalgorithm complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system performs preliminary learning during normal operation to establish baseline patterns of normal oil flow and particle distribution, enabling it to distinguish these from actual debris events and reduce false alarms before they occur

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The detection algorithm dynamically adjusts its parameters based on learned patterns from historical data, adapting threshold values and detection criteria to match actual operating conditions rather than relying on fixed manufacturer specifications

Inventive Principle:
Principle #35Parameter changes

3Productivity

If manual calibration is performed, then the manufacturing precision is adequate, but the productivity deteriorates due to time-consuming recalibration

Engineering Contradiction:
Improvemaintenance efficiencyVSAvoidcalibration precision
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The monitoring system performs self-calibration by automatically learning from operational data and adjusting its detection parameters without requiring manual intervention, eliminating time-consuming recalibration procedures while maintaining detection accuracy

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3312604B1Oil debris monitoring (ODM) with adaptive learning
Publication Date: 2022.09.28 RTX CORP
  • EP3312604B1 patent drawingFigure 1
  • EP3312604B1 patent drawingFigure 2
  • EP3312604B1 patent drawingFigure 3

AI summary

A method (700) for debris particle detection with adaptive learning includes: receiving (705) oil debris monitoring (ODM) sensor data from an oil debris monitor sensor and fleet data from a database; detecting (710) a feature in the ODM sensor data; generating an anomaly detection signal based on detecting (715) an anomaly by comparing the feature in the ODM sensor data to a limit defined by system information stored in the fleet data; selecting (720) a maintenance action request based on the anomaly detection signal; and adjusting one or more of the feature, the anomaly, the limit, and the maintenance action request by applying (725) an adaptive learning algorithm that uses the ODM sensor data, fleet data, and feedback from field maintenance of one or more engines that evolves over time.