Voice Assistant Privacy Mode via Secondary Device Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Personal electronic devices, such as smartphones, face issues with privacy when using voice commands, as audible responses can be inadvertently shared in public, potentially embarrassing the user if not in a private location or if not intended for them.

Innovation Solution

The implementation of an always-on privacy mode in user devices that discriminately provides audible responses by detecting secondary acknowledgments within a set time interval, triggering privacy mode and disabling loudspeaker playback if multiple devices are detected, and requiring user confirmation through audible or tactile inputs to verify authorized voice commands.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If voice commands are enabled with audible responses, then hands-free operation is improved, but privacy is compromised when multiple devices are present

Engineering Contradiction:
Improvehands-free operationVSAvoidprivacy exposure
Core Design Contradiction:
Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The system monitors for audible acknowledgments from other devices within a time interval after detecting a voice command. Based on this feedback about the presence of other devices, the system dynamically adjusts its response behavior - either providing audible responses or switching to private mode with tactile confirmation requirements

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system dynamically changes its operational mode based on real-time detection of other devices. It transitions between hands-free audible response mode and private mode with tactile confirmation, adapting its behavior to the current environmental context of device presence

Inventive Principle:
Principle #15Dynamics

2Ease of operation

If audible responses are provided for voice commands, then user convenience is improved, but information security deteriorates in public locations

Engineering Contradiction:
Improveuser convenienceVSAvoidinformation security
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The system uses feedback from monitoring other devices' acknowledgments to determine whether to provide audible responses. When other devices are detected, the system infers a public or shared environment and switches to private mode, preventing information leakage while maintaining convenience in private settings

Inventive Principle:
Principle #23Feedback

3Loss of information

If privacy mode is enforced to prevent information sharing, then information security is improved, but ease of operation deteriorates due to confirmation requirements

Engineering Contradiction:
Improveinformation securityVSAvoidoperation simplicity
Core Design Contradiction:
Loss of informationVSEase of operation

Solution Approach 1:

The system dynamically adjusts the level of confirmation required based on the detected environment. In private settings with no other devices, simple voice commands suffice. In public settings with other devices detected, the system requires additional tactile confirmation, adapting the operation complexity to the security needs of the situation

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3011426B1Privacy mode for always-on voice-activated information assistant
Publication Date: 2020.01.01 GOOGLE TECHNOLOGY HOLDINGS LLC
  • EP3011426B1 patent drawingFigure 1
  • EP3011426B1 patent drawingFigure 2
  • EP3011426B1 patent drawingFigure 3

AI summary

A user device and method discriminately provides audible responses to a voice command received by a user device that supports voice activation. The method includes detecting a first pre-established, audible activation command that activates the user device. In response to detecting the first pre-established, audible activation command, the method includes producing a first audible acknowledgement within loudspeaker proximity of the user device and then monitoring for detection of at least one second, audible acknowledgement produced by another user device within a pre-set time interval, which detection would indicate that the other user device is also responding. The method includes processing and responding to a received audible command in response to not detecting. However, in response to detecting, the method includes triggering entry into a privacy mode of audible command input and producing a privacy mode announcement via at least one of a display and a sound producing component.