Automated Event Detection and Duration Measurement for Industrial Assets
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial machines require efficient event detection and duration measurement without human supervision or training data, as existing methods are often noisy and costly, and lack automation in identifying asset states and optimizing performance.
Innovation Solution
A fully automated system using multiple sensors to detect events and states in industrial assets, creating mathematical models to identify asset types and optimize performance by selecting relevant sensors and determining event durations, which can be applied across various domains like aviation and transportation, without the need for training data or human supervision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If traditional event detection methods are used in industrial machines, then event detection capability is provided, but the system requires human supervision and training data, increasing operational complexity and cost
Solution Approach 1:
The system performs self-service by automatically detecting events and determining asset states without requiring human supervision or external training data. The unsupervised machine learning algorithms enable the system to autonomously learn from sensor data and identify events, eliminating the need for manual intervention in the detection process.
Solution Approach 2:
The patent replaces traditional mechanical or manual event detection systems with computational algorithms. By substituting human supervision and manual training data collection with unsupervised machine learning algorithms, the system achieves automated event detection while reducing operational complexity.
2Measurement precision
If multiple sensors are deployed to improve event detection accuracy, then measurement precision is improved, but system complexity and cost increase
Solution Approach 1:
The system performs preliminary action by pre-defining a finite set of possible asset states and transitions before actual event detection. This pre-specification of states and transition rules enables the system to process sensor data efficiently without requiring complex real-time analysis, thereby maintaining measurement precision while controlling system complexity.
Solution Approach 2:
The patent applies parameter changes by transforming raw sensor data into standardized asset state parameters through unsupervised learning. The system changes the parameter representation from raw sensor readings to meaningful asset states and transitions, improving detection accuracy while managing complexity through dimensionality reduction and pattern recognition.
3Ease of operation
If unsupervised learning algorithms are used to eliminate training data requirements, then ease of operation is improved, but measurement precision may be compromised
Solution Approach 1:
The system achieves universality by designing unsupervised learning algorithms that can detect multiple types of events and asset states across different industrial assets without requiring asset-specific training data. The same algorithmic framework universally applies to various asset types, maintaining ease of operation while achieving reliable event detection through pattern recognition in sensor data.
Solution Approach 2:
The system implements feedback mechanisms where the unsupervised learning algorithms continuously refine event detection based on observed sensor data patterns. The algorithms learn from the data stream itself, using feedback from the environment to improve detection accuracy over time without external training intervention, thus maintaining both ease of operation and measurement precision.
4Productivity
If asset states are automatically identified to optimize performance, then productivity is improved, but the difficulty of detecting and measuring states increases
Solution Approach 1:
The system applies segmentation by dividing continuous sensor data into discrete asset states and transitions. By segmenting the data stream into meaningful states based on predefined thresholds and patterns, the system enables automatic state identification that improves productivity while managing the complexity of detection through structured data organization.
Solution Approach 2:
The system performs preliminary action by pre-establishing the framework of possible asset states and transition rules before deployment. This pre-configuration enables automatic state identification during operation, improving productivity by eliminating manual state assessment while controlling measurement difficulty through the predefined state framework.
Data Source
AI summary
An asset class type of a new asset is predicted or determined based upon an evaluation of time series data from the new asset. A predicted asset type is used to identify sensors of the new asset to use for data collection. Using the readings of selected sensors from the new asset, states of the new asset are obtained. The duration at least one of these states of the new asset is determined. This information can be subsequently used to optimize the performance of the new asset.