Smart home company Habitat has officially launched its AI-native smart home architecture, the Gaia Architecture, during an evening event on September 8. Alongside this release, the company introduced multiple technological innovations, including the HClaw home agent, a home computing hub with space controllers, an AI-powered ultra-constant climate system, and the end-to-end AI design tool ANDI. Habitat announced it is moving beyond the traditional retrofit "component assembly" model of smart home technology, instead rebuilding residential spaces from the ground up to steer the housing industry into a new era of AI-native residences.

As AI agents rapidly integrate into the physical world, intelligence across various scenarios is accelerating, yet residential intelligence has stayed largely at a superficial level. According to Habitat's founder and CEO, Shen Yanan, smartphones represent a system-level overhaul centered on internet applications, and smart vehicles represent a similar overhaul around autonomous driving and intelligent cockpits. In the same vein, smart homes must undergo a system-level transformation built around an AI-native home agent. To achieve this, Habitat has outlined four distinct evolutionary phases for smart homes: starting with L1, the foundational stage combining integrated software and hardware with a home agent; progressing to L2, the advanced stage characterized by AI-native design and proactive services; advancing to L3, the high-end stage featuring embodied intelligence; and finally reaching L4, the ultimate stage of fully autonomous service within the smart home.

With the debut of the Gaia Architecture, Habitat's new generation L32 becomes the world's first AI-native L2-grade smart home product, officially marking the industry's transition into the L2 phase. Unlike conventional approaches that simply bolt AI features onto existing homes, the core principle of the Gaia Architecture is to treat the residence as a complete product developed through forward engineering. Every layer—from the underlying architecture and product design to business processes and commercial models—is constructed around AI reasoning, using integrated software and hardware to transform the living experience.

Homes equipped with the Gaia Architecture can actively understand household needs through perception and intelligent coordination. The system continuously expands functionality and improves performance through software OTA updates, decoupled from hardware constraints, minimizing disruption from complex devices. Residents in a Gaia-equipped home no longer need to learn how to operate complicated gadgets or coordinate between disconnected systems. Instead, the home proactively senses individual needs, adapts to users' states, and delivers seamless, intuitive living services.

Addressing common pain points in traditional smart home setups—where circuit panels are tightly bound to specific functions and future expansion is limited—Habitat has pioneered the application of automotive domain control architecture to residential settings, creating a more stable and flexible hardware foundation. By adopting a wired-backbone with wireless expansion approach, the system significantly boosts stability, reduces failure rates, and simplifies installation and maintenance. The space controllers support various power formats, including 24V low voltage, 220V mains electricity, and PoE Ethernet, while accommodating a wide array of device protocols such as infrared, Matter, Bluetooth, Wi-Fi, and 485 bus systems, providing a unified and reliable execution base for the home agent.

Furthermore, this architecture delivers comprehensive household energy management. On the hardware side, the environment system's energy module, smart breakers for high-power appliances, and energy interfaces on space controllers cover the collection of primary household electricity usage data. On the software side, the Xiaoqi assistant generates energy consumption reports, analyzes usage patterns, and offers optimization recommendations, making home energy monitoring and control fully transparent and actionable.