Smart Glasses Chat Architecture Using Local and Server AI Models

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing smart glasses have limited functionality and low intelligence, primarily serving as audio devices with basic functions like listening to music and making calls, lacking advanced interaction and intelligence.

Innovation Solution

Integration of Generative Artificial Intelligence Large Language Models (GAILLM) into smart glasses, enabling chat functions through a microphone, speaker, processor, and memory, with connectivity to a smart mobile terminal and model server for converting user speech into intelligent replies.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If Generative AI Large Language Model is integrated into smart glasses, then intelligence and interactivity are improved, but device complexity increases

Engineering Contradiction:
ImproveintelligenceVSAvoiddevice complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The patent introduces a server as an intermediary component that hosts the large language model. The smart glasses contain only a local language model, and the server acts as a mediator to process complex language tasks. This allows the glasses to provide intelligent chat functions without embedding the entire large language model locally, thus improving intelligence while controlling device complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system is divided into multiple segments: the smart glasses contain a local language model for basic processing, while the server contains the comprehensive large language model for advanced processing. This segmentation allows the intelligent functions to be distributed across devices, with the glasses handling user interaction and the server handling complex computation.

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If chat function with GAILLM is added to smart glasses, then functionality is improved, but device complexity increases

Engineering Contradiction:
ImprovefunctionalityVSAvoiddevice complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The server is designed as a universal platform that can provide various language model services to multiple smart glasses devices. The local language model in the glasses provides basic chat functionality, while the server's large language model enhances this with advanced capabilities. This multi-functionality approach allows the system to provide diverse functions without adding complexity to each individual glass device.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Speed

If local language model is embedded in smart glasses, then processing speed is improved, but manufacturing cost increases

Engineering Contradiction:
Improveprocessing speedVSAvoidmanufacturing cost
Core Design Contradiction:
SpeedVSEase of manufacture

Solution Approach 1:

Instead of embedding the complete large language model in the smart glasses, the patent implements a partial language model locally that handles basic processing quickly. More complex processing is offloaded to the server. This partial action approach provides sufficient processing speed for everyday use while avoiding the high manufacturing costs of embedding the full model.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4715446A1Smart glasses based on generative artificial intelligence large language model, system and control method
Publication Date: 2026.03.25 SOLOS TECH SHENZHEN LTD
  • EP4715446A1 patent drawingFigure 1
  • EP4715446A1 patent drawingFigure 2
  • EP4715446A1 patent drawingFigure 3

AI summary

Smart glasses, a control method and a system based on generative artificial intelligence large language models, wherein the smart glasses include a front frame, a temple, a microphone, a speaker, a processor and a memory, one or more computer programs executable on the processor are stored in the memory, the one or more computer programs include instructions for: activating a chat function of the smart glasses in response to a first control instruction; obtaining a first speech of a user through the microphone, wherein the first speech includes a question asked by the user; obtaining a second speech including a reply corresponding to the question through the generative artificial intelligence large language models, and playing the second speech through the speaker. The application improves the intelligence and interactivity of the smart glasses.