# A New Kind of Furniture Is Learning How to Respond
For decades, furniture has relied on human memory to work properly. Drawers, tables, cabinets, and storage boxes all assume that people will remember where things were placed—even as homes become more complex and storage spreads across rooms and surfaces. The result is a quiet, daily friction: time lost searching for objects that already exist somewhere nearby.

A robotics startup based in Cupertino is quietly working on a system designed to remove that burden.


LabartCasa Robotics has been developing a novel AI-powered storage and furniture platform that allows people to store everyday items and retrieve them using natural, conversational requests. The system is designed to function as—or integrate into—common furniture forms such as drawers, side tables, and compact storage units, blurring the line between furniture and intelligent systems.

Instead of searching through drawers or relying on memory, users interact with the furniture by simply speaking—turning physical storage into something closer to an assistant than a container.

The technology relies on a combination of automatic object recognition and semantic retrieval. Items are visually recognized as they are placed into storage and indexed based on descriptive attributes rather than fixed labels. When a user makes a request like “Where is my wallet?” or “Give me something sharp,” the system interprets intent and meaning, matching the request to previously observed physical items.

The company has filed multiple patents across a broader patent family covering key aspects of the system, including recognition-driven storage, semantic matching between natural language and physical objects, and interaction models that allow furniture to manage recall autonomously. The technology is currently protected under those filings as development continues.

“We’re not trying to build a gadget that feels technical,” one of the founders said. “The goal is furniture that feels natural to live with—something you talk to the same way you’d ask another person for help.”

While the company has not yet shared product images or final specifications, it confirmed that the first version is being designed specifically for home use, with a focus on compact form factors and simple, intuitive interaction. The intent, according to the team, is to make the intelligence feel em

LabartCasa Robotics also noted that it is considering a future early-access release, potentially limited in scale to gather feedback from first users. People interested in following development or receiving updates can leave their email on the company’s website:

[labartcasa.github.io](https://labartcasa.com/)