Edge Early Warning Using Confidence Score Deviations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional early warning systems for edge devices are not sensitive or robust enough to predict hardware or software failures confidently due to inadequate data availability or usage, leading to potential disruptions.
Innovation Solution
Implementing a DCF-assisted early warning system that monitors data confidence scores to identify potential issues in edge devices by triggering alerts or initiating corrective actions when confidence score deviations exceed predefined thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional early warning systems are used to monitor edge devices, then system simplicity is maintained, but prediction sensitivity and reliability are insufficient due to inadequate data availability and usage
Solution Approach 1:
The patent introduces a data confidence fabric (DCF) as an intermediary layer between edge devices and the monitoring system. The DCF generates confidence scores that mediate the relationship between raw device data and failure predictions, enabling reliable predictions without requiring complex analysis of raw data from multiple sources
Solution Approach 2:
The patent replaces traditional mechanical/data-intensive monitoring approaches with a confidence score-based system. Instead of analyzing raw sensor data, device logs, and performance metrics directly, the system uses DCF-generated confidence scores as a substitute metric that encapsulates failure risk information
2Measurement precision
If more data types and amounts are collected to improve failure prediction accuracy, then prediction confidence increases, but data processing complexity and resource consumption increase
Solution Approach 1:
The patent extracts the essential failure prediction information from complex multi-source data into a single confidence score metric. The DCF system takes out the critical signal from noisy, diverse data sources and presents it in a simplified form that can be monitored without processing the underlying complexity
Solution Approach 2:
The patent transforms multiple data parameters (sensor readings, log entries, performance metrics) into a single confidence score parameter. This parameter transformation simplifies monitoring while maintaining prediction accuracy, as the confidence score encapsulates the state of all underlying parameters
Data Source
AI summary
One example method includes monitoring confidence scores received from an edge entity in an edge environment, and the edge entity comprises hardware and/or software, comparing a change in the confidence scores with a threshold permissible confidence score change, when a magnitude of the change is above a threshold permissible change, identifying the edge entity, or a system/device monitored by the edge entity, as a possibly malfunctioning entity, identifying a corrective action for the edge entity, and implementing, and/or causing implementation of, the corrective action.