Terminal AI Model Management for Compatibility and Privacy

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing user terminal environments face limitations in integrating large AI models due to memory and inference-specific chip constraints, and there are concerns about privacy exposure and network delays with external AI models, necessitating a secure and efficient management system for AI models operating on terminals.

Innovation Solution

A method and system for managing AI models on terminals involve storing device information, determining compatibility, setting access ranges, and securely installing and updating models, including encryption and secure data access management, to ensure safe and efficient operation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If large AI models are deployed on terminals, then AI performance is improved, but terminal memory and hardware resources are insufficient

Engineering Contradiction:

ImproveAI performanceVSAvoidmodel size

Core Design Contradiction:

ReliabilityVSWeight of moving object

Solution Approach 1:

The patent segments AI models into different size categories (first AI models and second AI models) based on their characteristics. The system selectively deploys appropriate models to terminals based on terminal capabilities, dividing the model space to match terminal hardware constraints while maintaining high performance through specialized smaller models.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system changes the parameter of model size selection based on terminal characteristics. By adjusting which models are deployed to which terminals according to hardware capabilities, the system optimizes the balance between performance and resource consumption, enabling high AI performance within terminal memory constraints.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If AI models operate on external networks, then functionality is provided, but privacy is exposed and network delays occur

Engineering Contradiction:

ImproveAI functionalityVSAvoidprivacy exposure and network delay

Core Design Contradiction:

Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The patent introduces a model management server as an intermediary between AI model providers and terminals. This server mediates the deployment process by verifying terminal eligibility, selecting appropriate models, and managing the distribution, thereby enabling secure AI functionality while controlling privacy exposure through centralized management.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system creates copies of AI models that are deployed to specific terminals based on their characteristics. By making model copies available to eligible terminals, the system enables AI functionality locally without requiring continuous external network access, thus reducing network delays and privacy exposure risks.

Inventive Principle:

Principle #26Copying

3Adaptability or versatility

If multiple AI models are managed on terminals, then functionality is enhanced, but device complexity increases

Engineering Contradiction:

ImproveAI model functionalityVSAvoidmodel management complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements a universal model management server that handles multiple AI models with different characteristics through a single centralized system. This multi-functional server manages model verification, selection, and deployment for various terminal types, reducing the complexity that would otherwise exist in managing multiple models across different devices.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system incorporates feedback mechanisms where the model management server receives information about terminal characteristics and model performance, then uses this feedback to optimize future model selections and deployments. This feedback loop simplifies management by automatically adapting to changing conditions without manual intervention.

Inventive Principle:

Principle #23Feedback

4Loss of energy

If AI models are selectively deployed based on terminal characteristics, then resource efficiency is improved, but compatibility determination complexity increases

Engineering Contradiction:

Improveresource efficiencyVSAvoidcompatibility determination complexity

Core Design Contradiction:

Loss of energyVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-verifying terminal eligibility and pre-determining model compatibility before actual deployment. The model management server performs these compatibility determinations in advance based on stored terminal characteristics, so that when deployment time arrives, the process is simplified and resource-efficient without complex real-time analysis.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS20250383861A1Method and system for managing artificial intelligence model installed and operating in terminal environment

Publication Date: 2025.12.18 PIAMOND CORP

AI summary

Method and system for managing artificial intelligence model installed and operating in terminal environment are disclosed. A method for managing an artificial intelligence model according to one embodiment may include storing and updating device information of each of a plurality of user terminals that download and install an artificial intelligence model from an artificial intelligence model store, and determining whether or not a first artificial intelligence model can be installed on a first user terminal based on device information of a first user terminal according to an installation request for the first artificial intelligence model of the first user terminal among the plurality of user terminals.