How Sonos Speakers Use Personalized Wake Words to Recognize Different Users

A system that lets multiple people control a shared speaker by using unique voice-trigger words to link their specific music accounts and preferences.

What it covers

This patent describes a way for a smart speaker to distinguish between different people in a home by assigning each person a unique 'wakeup word.' When a user speaks their specific word, the system identifies their linked user profile and music service account. It then executes commands—like playing a playlist or adjusting volume—using that specific person's settings and account history. For example, if you say 'Alexa, play jazz' and your partner says 'Hey Sonos, play rock,' the system knows exactly which music account to pull from based on the unique trigger word used.

What it doesn't cover

• —Does not cover voice recognition based on biometric voice printing or vocal characteristics.
• —Does not cover systems that rely on a single, universal wake word for all users.
• —Does not cover controlling media devices that are not connected to a remote music service account.

The clever bit

The system uses the wake word itself as a routing mechanism to switch between different cloud-based music service accounts, effectively turning a single hardware device into a multi-user portal without needing complex login screens.

Why it matters

This technology addresses the 'shared device' problem in smart homes. By using wake words as a proxy for user identity, it allows households to maintain separate music libraries and preferences on a single piece of hardware, preventing one user's listening habits from polluting another's recommendations.

Real-world examples

• 1.Sonos smart speakers with integrated voice control
• 2.Multi-user smart home audio environments