Machine Learning Equipment Damage Prediction From Sensor Clusters

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current predictive maintenance strategies for equipment with moving parts face challenges in determining reliable threshold levels for warnings and alarms, leading to inefficiencies in detecting equipment deterioration and preventing downtime.

Innovation Solution

A data-driven approach using machine learning algorithms to analyze sensor data and determine equipment states by identifying clusters representative of different equipment conditions, without requiring prior knowledge or pre-programmed information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional threshold-based monitoring is used, then equipment failure can be detected, but reliable threshold levels for warnings and alarms are difficult to determine

Engineering Contradiction:
Improvethreshold determination accuracyVSAvoidthreshold setting complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system automatically determines threshold levels by analyzing historical sensor data and identifying clusters representing different equipment states. The machine learning model self-adjusts thresholds based on learned patterns from calibration data, eliminating the need for manual threshold configuration by experts.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system transforms fixed threshold values into dynamic, data-driven parameters. By using clustering algorithms on historical sensor data, the system identifies state boundaries and automatically adjusts threshold levels based on actual equipment behavior patterns rather than static predetermined values.

Inventive Principle:
Principle #35Parameter changes

2Productivity

If predictive maintenance is implemented, then downtime can be reduced, but accurate prediction of equipment deterioration requires reliable threshold values

Engineering Contradiction:
Improveequipment availabilityVSAvoidprediction accuracy
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary analysis by collecting and analyzing calibration sensor data before actual predictive maintenance is needed. Clusters representing normal and abnormal equipment states are identified in advance, establishing reliable thresholds that enable accurate future predictions without requiring threshold adjustment at the time of prediction.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses historical sensor data as feedback to continuously improve threshold determination. By analyzing past equipment behavior patterns and cluster distributions, the system refines its understanding of normal versus abnormal states, thereby improving prediction accuracy for future maintenance decisions.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If machine learning algorithms are used to analyze sensor data, then equipment state prediction accuracy is improved, but the system requires calibration data and training

Engineering Contradiction:
Improveequipment state prediction accuracyVSAvoidsystem implementation complexity
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The system performs preliminary training using calibration sensor data collected during equipment installation or commissioning. This upfront training phase establishes the clustering model and threshold levels, after which the system can operate autonomously without requiring additional manual configuration or retraining.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3712728B1Apparatus for predicting equipment damage
Publication Date: 2025.05.14 ABB (SCHWEIZ) AG
  • EP3712728B1 patent drawingFigure 1~2
  • EP3712728B1 patent drawingFigure 3
  • EP3712728B1 patent drawing

AI summary

The present invention relates to an apparatus for predicting equipment damage. The apparatus comprises an input unit, a processing unit, and an output unit. The input unit is configured to provide the processing unit with sensor data for an item of equipment. The processing unit is configured to implement at least one machine learning algorithm. The at least one machine learning algorithm has been trained on the basis of a plurality of calibration sensor data for the item of equipment. Training of the at least one machine learning algorithm comprises processing the plurality of calibration sensor data to determine at least two clusters representative of different equipment states. The processing unit is configured to implement the at least one machine learning algorithm to process the sensor data to assign the sensor data to a cluster of the at least two clusters to determine an equipment state for the item of equipment. The output unit is configured to output the equipment state for the item of equipment.