Abu Dhabi-based Origen unveiled DOMIA at Mobile World Congress in Barcelona — a smart home platform that replaces the typical “if-this-then-that” routines with LLM-powered AI agents that can understand context, learn daily habits, and coordinate actions across an entire living space.
The system’s most notable design choice: it deliberately avoids cameras. Instead, DOMIA uses mm-wave radar and LiDAR for presence detection, paired with a digital twin that provides a live 3D view of the home for monitoring and control.
Table of Contents
How it works
DOMIA runs a multi-agent AI architecture where specialized agents work together to handle complex requests. Say “it’s a bit dark” and the system adjusts lighting differently depending on time of day and what you’re doing. Ask it to “prepare the house for guests” and it coordinates comfort settings, lighting, and access permissions dynamically — not from a preset mode.
The core processing happens locally on what Origen calls an “AI Box” — an edge-native device that runs reasoning and decision-making inside the home rather than in the cloud. All data stays on-device, which addresses both latency and privacy concerns in one move.
Why it matters
Smart home platforms have been stuck in a rut of rigid automations and cloud-dependent voice assistants for years. DOMIA’s approach of bringing LLM reasoning to the edge — with genuine contextual understanding rather than keyword matching — represents the kind of leap the industry has been promising but rarely delivering. The camera-free privacy stance could also be a meaningful differentiator as smart home adoption pushes into more privacy-conscious markets.
Origen is an Abu Dhabi technology company specializing in sovereign digital infrastructure for AI systems, smart manufacturing, and urban intelligence.
