Fleet Thermal Grading Using Anomaly Models for Device Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices often experience thermal issues that can lead to damage or failure, and existing methods lack effective predictive and corrective measures to manage these issues across a fleet of devices with varying thermal characteristics.
Innovation Solution
A fleet management system collects thermal data and applies anomaly models to calculate thermal grades for devices and their components, identifying corrective actions to improve thermal performance, including scheduling maintenance and repairs, and controlling devices to perform these actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If thermal monitoring and anomaly detection systems are implemented across a fleet of devices, then device reliability and lifespan are improved, but system complexity and computational requirements increase
Solution Approach 1:
The system segments the fleet management into multiple anomaly models, each dedicated to detecting specific thermal issues. Instead of one monolithic complex system, multiple specialized models divide the detection task, reducing individual model complexity while maintaining overall reliability through comprehensive coverage of different failure modes.
Solution Approach 2:
The patent introduces a fleet management system as an intermediary layer between individual devices and the anomaly detection process. This intermediary aggregates thermal data from multiple devices, applies standardized anomaly models, and produces centralized thermal grades, simplifying the architecture by separating data collection from analysis and enabling reusable detection logic.
2Measurement precision
If multiple anomaly models are applied to thermal data to produce comprehensive thermal grades, then measurement precision and diagnostic accuracy are improved, but computational time and processing requirements increase
Solution Approach 1:
The system performs preliminary actions by collecting and storing thermal data over time before actual anomaly detection is needed. Thermal data is accumulated from multiple devices and time points, preparing the dataset in advance so that when anomaly models need to be applied, the processing can proceed more efficiently with pre-organized data rather than collecting it during the analysis phase.
Solution Approach 2:
Multiple anomaly models are merged into a unified fleet management system that processes thermal data collectively. Instead of running separate detection systems independently, the patent combines multiple models to work together on the same thermal dataset, producing integrated thermal grades that leverage synergies between different detection approaches while sharing computational resources.
Data Source
AI summary
An example of a computer-readable medium storing machine-readable instructions. The instructions may cause the processor to receive thermal data for a device and apply anomaly models to the thermal data to produce grades. Grades for a device may be combined into a system thermal grade and corrective actions identified to improve the system thermal grade.