Dynamic Persona Assignment for Edge Device Optimization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing information handling systems face challenges in dynamic persona assignment for optimizing near-end and edge devices, as static persona detection is inadequate for real-time adaptive experiences and cloud-based persona detection faces data movement issues such as security, latency, and cost.

Innovation Solution

An information handling system that includes a processor capable of determining telemetry data types, selecting appropriate machine learning models, and assigning a persona based on constraints, allowing for dynamic persona assignment that can be computed on the information handling system, edge device, or cloud server.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If cloud-based persona detection is used, then comprehensive data analysis is possible, but data movement issues such as security, latency, and cost increase

Engineering Contradiction:

Improvepersona detection accuracyVSAvoiddata movement latency

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent implements federated learning where machine learning models are executed locally on edge devices and information handling systems rather than centralizing all data processing in the cloud. This allows persona detection to occur at the source of data generation, eliminating data movement latency while maintaining detection accuracy through distributed model execution across multiple locations.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system segments the persona detection process into distributed components: telemetry data collection occurs locally at edge devices, machine learning model execution is distributed across multiple nodes, and only aggregated insights are shared. This segmentation enables parallel processing and eliminates the need to move large volumes of raw data to centralized cloud servers.

Inventive Principle:

Principle #1Segmentation

2Adaptability or versatility

If static persona detection is used, then system complexity is reduced, but real-time adaptive experiences cannot be provided

Engineering Contradiction:

Improvereal-time adaptive capabilityVSAvoiddynamic persona assignment complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements dynamic persona assignment where machine learning models continuously analyze telemetry data and update persona assignments in real-time based on changing user behavior and system conditions. This dynamic approach enables real-time adaptive experiences while the system manages complexity through automated model execution and constraint-based decision making.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system employs self-service mechanisms where edge devices and information handling systems autonomously execute machine learning models and perform persona detection locally without requiring constant centralized management. This self-service capability reduces operational complexity while enabling real-time adaptability through autonomous decision-making at the edge.

Inventive Principle:

Principle #25Self-service

3Productivity

If machine learning models are executed on edge devices, then real-time persona detection is achieved, but device resource constraints are challenged

Engineering Contradiction:

Improvepersona detection speedVSAvoidedge device energy consumption

Core Design Contradiction:

ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent implements a hybrid execution strategy where machine learning models are selectively executed on edge devices only when necessary for real-time persona detection, while less time-sensitive processing can occur elsewhere. This partial execution approach balances the need for real-time detection speed with the constraints of edge device energy and computational resources.

Inventive Principle:

Principle #16Partial or excessive action

4Reliability

If federated learning is implemented, then data security is improved, but system coordination complexity increases

Engineering Contradiction:

Improvedata securityVSAvoidfederated learning coordination complexity

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a universal constraint management framework that handles multiple concerns (security, latency, cost, energy) through a single standardized interface. This universal approach simplifies the coordination complexity of federated learning by providing a unified mechanism for managing distributed machine learning model execution across diverse edge devices and cloud environments.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS20250139463A1Dynamic persona assignment for optimization of near end and edge devices

Publication Date: 2025.05.01 DELL PROD LP

AI summary

An information handling system stores telemetry data. A processor determines one or more types of the telemetry data. The processor determines one or more machine learning models to be executed. Each different machine learning model corresponds to a different type of telemetry data. The processor determines one or more constraints for the machine learning models. Based on the one or more constraints, the processor determines a device to execute the machine learning models. The processor executes the machine learning models in the determined device. The telemetry data is provided as inputs to the machine learning models. Based on the execution of the machine learning model, the processor determines a persona for the information handling system.