Radar Enabled Keyword Activation for Voice Assistants

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Voice assistants require a predefined wakeup keyword for command recognition, limiting natural conversation and interaction, as users desire a more intuitive way to engage with the system without the need for explicit keyword activation.

Innovation Solution

A system utilizing radar technology, specifically frequency modulated continuous wave (FMCW) radar, to detect and locate individuals within a room, allowing for context-aware keyword activation based on presence, location, and predefined segments, enabling voice assistant interaction without the need for a global keyword in specific situations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a predefined wakeup keyword is required for voice assistant activation, then command recognition accuracy is improved, but natural conversation flow deteriorates

Engineering Contradiction:
Improvecommand recognition accuracyVSAvoidnatural conversation flow
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The radar system performs preliminary detection of user presence and location before voice processing begins. This preliminary action allows the system to pre-configure activation parameters based on detected user context, enabling natural conversation without requiring explicit wakeup keywords when users are properly positioned.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system dynamically changes the activation parameters of the voice assistant based on radar-detected user presence and location. When a user is detected in a predefined segment, the system modifies the activation threshold or keyword requirement parameters, allowing keyword-free activation in appropriate contexts while maintaining reliability when needed.

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If radar technology is implemented for context-aware activation, then natural interaction is improved, but device complexity increases

Engineering Contradiction:
Improvenatural interactionVSAvoidsystem architecture
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The radar system serves as an intermediary component that bridges physical presence detection and voice processing. This intermediary provides context information (user presence, location, movement) that enables intelligent activation decisions without requiring complex integration between voice and other sensors, simplifying the overall system architecture.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system divides the operational space into predefined segments and processes activation logic separately for each segment. This segmentation allows the complex activation logic to be distributed and managed in modular fashion, reducing overall system complexity by breaking down the decision-making process into manageable regional zones.

Inventive Principle:
Principle #1Segmentation

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enables a more natural and intuitive interaction with voice assistants by allowing command execution based on user presence and location, reducing the reliance on predefined keywords and enhancing user experience through context-aware processing.

Implementation Method 1

A system utilizing radar technology, specifically frequency modulated continuous wave (FMCW) radar, to detect and locate individuals within a room

Methodology Applied
Scientific EffectRadar: Radar

Data Source

PatentEP3547309B1Radar enabled location based keyword activation for voice assistants
Publication Date: 2024.02.28 INFINEON TECHNOLOGIES AG
  • EP3547309B1 patent drawingFigure 1
  • EP3547309B1 patent drawingFigure 2
  • EP3547309B1 patent drawingFigure 3

AI summary

A method of operation for a voice assistant includes detecting the presence of a person and sound in a general location, detecting whether the person is in a specific location of the general location, and if the person is in the specific location, initiating operation of the voice assistant with a predefined specific command other than a global keyword.