Sensor Event Correlation for False-Alarm Reduction in Asset Monitoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In industrial settings, predictive maintenance and failure detection are hindered by the complexity of systems and the generation of false alarms from machine-learning algorithms, making it difficult to identify significant changes in sensor combinations and determine the underlying causes of events.
Innovation Solution
A method and system for improving asset operation by creating an interconnected representation of a complex physical operation, including an asset representation with sensor listings, object listings, process listings, and entity connections, which generates calculated indicators and hints to highlight impairing factors, using machine-learning algorithms and user feedback to prioritize and modify hints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If machine-learning algorithms are used to analyze sensor data, then valuable information can be generated, but false alarms increase and analysis complexity increases
Solution Approach 1:
The patent segments the complex sensor data analysis into distinct components: individual sensor monitoring, event detection, and hierarchical relationship analysis. By breaking down the analysis into manageable segments and processing them separately, the system reduces false alarms while maintaining valuable information extraction capabilities.
Solution Approach 2:
The patent applies local quality by focusing analysis on specific local patterns and relationships within sensor data rather than treating all data uniformly. It identifies and analyzes local events and their hierarchical relationships, allowing the system to distinguish between significant local changes and normal variations, thereby reducing false alarms.
2Loss of information
If machine-learning algorithms are used to analyze sensor data, then valuable information can be generated, but the complexity of analysis increases
Solution Approach 1:
The system segments the analysis process into distinct hierarchical levels: individual sensor analysis, event detection, and relationship analysis. This segmentation reduces overall complexity by allowing each segment to be processed independently with appropriate algorithms, rather than requiring a single complex analysis system.
Solution Approach 2:
The patent introduces a hierarchical dimension to the analysis by organizing events into parent-child relationships and analyzing them across multiple levels. This dimensional approach transforms the complex multidimensional sensor data into a structured hierarchy that is easier to analyze and interpret while preserving valuable information.
3Measurement precision
If comprehensive sensor monitoring is implemented, then asset performance can be monitored, but false alarms and difficulty in identifying significant changes increase
Solution Approach 1:
The patent applies local quality by focusing on specific local patterns and relationships within the comprehensive sensor data. Instead of treating all sensor changes equally, it identifies and analyzes local events with their specific characteristics and hierarchical relationships, allowing the system to distinguish significant changes from normal variations even in comprehensive monitoring.
Solution Approach 2:
The patent adds a hierarchical dimension to comprehensive sensor monitoring by organizing sensor events into parent-child relationships. This dimensional transformation allows the system to analyze events at multiple levels of abstraction, improving the ability to identify significant changes while filtering out noise from comprehensive monitoring.
Data Source
AI summary
The present invention discloses methods and systems for improving asset operation based on identification of significant changes in sensor combinations in related events. Methods include the steps of: providing an asset representation of an asset having an alert-aggregation system; creating a calculated indicator listing having at least one calculated indicator in the asset configured to calculate, predict, or estimate an indicated asset behavior; incorporating the calculated indicator listing into a set of entity connections by associating listing elements in the calculated indicator listing, thereby producing the interconnected representation; associating the identified asset behavior with at least one relevant object and/or at least one relevant process that is impairing the asset from optimally performing, conducting, and/or achieving the identified asset behavior; and generating at least one hint associated with at least one relevant object and/or at least one relevant process that is impairing the identified asset behavior.