Home Appliance Voice Recognition Server Offloading

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing voice recognition systems for home appliances face limitations in recognizing natural language and processing voice commands efficiently, particularly due to resource constraints in individual apparatuses, leading to high computation requirements and increased probability of false recognition, which affects user convenience and reliability.

Innovation Solution

A method and system that utilize a voice recognition server system to process voice commands, offloading the natural language processing from the home appliance to a network of servers, allowing for efficient recognition and minimizing false recognition by using Automatic Speech Recognition (ASR), Natural Language Processing (NLP), and Text-to-Speech (TTS) servers, and providing audible guidance for user feedback.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If voice recognition is implemented using embedded modules in individual home appliances, then the system can process voice commands locally, but the computation requirements exceed the available system resources

Engineering Contradiction:
Improvevoice recognition accuracyVSAvoidsystem resource requirements
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces a voice recognition server as an intermediary component that handles the complex natural language processing tasks. The home appliance contains only a simple voice recognition module that captures voice input and transmits it to the server, which then processes the speech-to-text conversion and natural language understanding. This mediator approach allows the appliance to achieve high recognition accuracy without requiring substantial embedded computational resources.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If simple keyword recognition is used in the home appliance, then the system resources are conserved, but the ability to recognize natural language is limited

Engineering Contradiction:
Improvenatural language recognition capabilityVSAvoidcomputation power
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The voice recognition system is segmented into two distinct functional parts: a lightweight voice capture and transmission module embedded in the home appliance, and a sophisticated speech-to-text and natural language processing module hosted on the voice recognition server. This segmentation allows the appliance to maintain simplicity and low resource consumption while the server handles the computationally intensive tasks of recognizing and interpreting natural language commands.

Inventive Principle:
Principle #1Segmentation

3Reliability

If voice recognition processing is performed locally in the home appliance, then response time may be faster, but the probability of false recognition increases due to limited processing capability

Engineering Contradiction:
Improverecognition accuracyVSAvoidvoice command processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements a voice recognition server that maintains speech-to-text conversion capabilities as a centralized service. Multiple home appliances can connect to this server and utilize its processing power without each appliance needing to replicate the complex recognition algorithms. This copying approach allows appliances to access high-accuracy recognition capabilities on-demand while minimizing local resource consumption and reducing false recognition rates.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3404655B1Home appliance and method for operating the same
Publication Date: 2020.09.23 LG ELECTRONICS INC
  • EP3404655B1 patent drawingFigure 1
  • EP3404655B1 patent drawingFigure 2
  • EP3404655B1 patent drawingFigure 3a

AI summary

Disclosed is a method for operating a home appliance, including: setting a wake-up signal standby mode for receiving a wake-up voice signal comprising a preset call word; receiving a voice signal via an audio input unit comprising at least one microphone; when the voice signal comprises the call word, transmitting the received voice signal to the voice recognition server system; receiving analysis result information of the voice signal from the voice recognition server system; and, when the received analysis result information comprises information indicating that the voice signal comprises the call word, setting a command standby mode for receiving a user's voice command input for a preset period of time.