Edge Machine Learning Model Orchestration for Simultaneous Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Edge computers, with limited resources, face challenges in maintaining model accuracy and compatibility due to changing subjects or situations, leading to delayed and manual model replacements that cause downtime and inefficiencies.
Innovation Solution
A system and method for automatically detecting changes on edge computers, grouping similar devices, and selecting optimal models for simultaneous updates across multiple edge computers based on correlations and historical data analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If manual model replacement is used on edge computers, then device complexity is reduced, but loss of time increases due to delays in identifying and invoking replacement procedures
Solution Approach 1:
The system enables edge computers to automatically detect when model replacement is needed and invoke the replacement process without manual intervention. The edge computer monitors its own performance and environment, identifies when a model update is required, and autonomously initiates the replacement procedure, eliminating delays associated with manual detection and invocation.
Solution Approach 2:
The system implements feedback mechanisms where edge computers continuously monitor their operational performance and environmental conditions. When performance degradation or environmental changes indicate the need for model replacement, the system receives feedback and automatically triggers the replacement process, creating a closed-loop system that responds to conditions without human intervention.
2Productivity
If simultaneous model replacement on multiple edge computers is implemented, then productivity is improved by reducing downtime, but device complexity increases due to orchestration requirements
Solution Approach 1:
The system merges the model replacement operations of multiple edge computers into a coordinated simultaneous update process. By grouping edge computers and synchronizing their model replacement operations, the system achieves efficient batch updates that reduce overall downtime while managing complexity through unified orchestration of multiple devices.
Solution Approach 2:
The system performs preliminary actions by pre-configuring replacement models and preparing edge computers for simultaneous updates. Edge computers are pre-grouped and pre-configured with replacement models before the actual update event, allowing coordinated replacement to proceed efficiently without last-minute complexity management.
3Ease of operation
If human operators manually identify and invoke model replacement, then ease of operation is maintained, but loss of time increases due to lengthy delays in detection and action
Solution Approach 1:
The system transfers the manual operation of model replacement to automated self-service functionality. Edge computers automatically monitor their own operational status and environment, detect when model replacement is needed, and invoke the replacement process autonomously, eliminating the time delays associated with human operator response while maintaining operational simplicity through automated decision-making.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Example implementations described herein involve systems and methods to substantially simultaneously orchestrate machine learning models over multiple resource constrained control edge devices, so that the overall system is more agile to changes in events and environmental conditions where the models have been deployed. The example implementations described herein involve multiple processes that when executed, determine a list of edge devices to be updated along with the corresponding models based on correlation.