Point and clean: Matic adds gesture and voice control to its robot vacuum

The new features, rolling out today, leverage the robot’s existing array of five cameras to track user gestures. By saying “Hey, Matic” and pointing at a spill, the vacuum identifies the target area and begins a spot clean. Unlike traditional app-based zone drawing, which often lacks precision, this vision-based system allows for more granular control. Testing reveals that the robot responds accurately to height-dependent gestures—crouching down results in a targeted spot clean, while standing gestures trigger a wider cleaning radius.

Voice capabilities have also seen a significant boost. Powered by Google’s Gemini API, the assistant moves beyond rigid, pre-programmed commands. Users can issue conversational instructions like “vacuum the inside of the carpet” or “clean the kitchen, then mop the bathroom.” While the “Hey, Matic” wake word is processed locally, subsequent commands are routed to the cloud to support broader language flexibility. For privacy-conscious users, Matic allows the entire Cues system and the microphones to be disabled via the app settings.

This update arrives just before a scheduled price increase on September 9, when the hardware cost will rise from $1,245 to $1,495. Founded by former Google Nest engineers who have been refining this computer vision technology since 2017, the company intends to transition the most frequently used voice commands to on-device processing in future updates, aiming to bridge the gap between cloud-enabled convenience and local privacy.