Xiaomi Open-Sources Robotics Foundation Model Trained on 100K Hours of Trajectories
CyberRobooo · x · 2026-08-05
Xiaomi has released Xiaomi-Robotics-1, an open-source foundation model for robots designed to break the data bottleneck in robotics. The model combines large-scale embodiment-free pre-training with a post-training stage using real-robot data.
Data & Collection
- Pre-training: Utilizes 100,000 hours of embodiment-free (UMI) trajectories across 1,700+ scenarios (household, commercial, industrial, outdoor). A scalable auto-labeling pipeline processes state transitions.
- Post-training: Incorporates cross-embodiment datasets including in-house real-robot data (7,200+ hours of home tasks like tidying sofas), filtered open-sourced data, and high-quality UMI data.
Methodology
Following the LLM training paradigm, the project investigates the scaling laws of robot policy models concerning data, parameters, and compute.
Related event: Xiaomi Open-Sources Robotics Foundation Model Robotics-1(4 posts)→
More from Embodied
- DOBOT Unveils LUMO: A 1.3-Meter Home Humanoid Focused on Social Intelligence — ChrisGPT · 2026-08-06
- FCC Advanced Robot Ban: Devices with >35% Foreign Components Face Prohibition — mattfreed · 2026-08-06
- Physical Robot Chess Champions Outperform Humans Like Stockfish Beating Bots — srchvrs · 2026-08-06
- Embodied AI in Action: Robot Autonomously Collects Multi-Lighting Image Data Overnight — alexisgallagher · 2026-08-05
- Ginkgo to Build 4 New Autonomous Labs at MIT, Caltech, and More — ProfBuehlerMIT · 2026-08-05
- Jensen Huang on Digital Immortality: Sending His Humanoid Robot to Space — r0ck3t23 · 2026-08-05