RoboOS

First open-source embodied operating system based on a Brain-Cerebellum hierarchical architecture for cross-embodiment and multi-agent collaboration. Three components: (1) Embodied Cloud Model (MLLM for global perception), (2) Cerebellum Skill Library (plug-and-play toolkit), (3) Real-Time Shared Memory (spatiotemporal synchronization). Supports heterogeneous robots including single-arm, dual-arm, humanoid, and wheeled robots. RoboOS 2.0 adds MCP-powered SaaS framework with RoboSkill Store. Tested in restaurants, households, and supermarkets.

Unified memory-based framework for lifelong, scalable, and robust multi-robot collaboration. Introduces Spatio-Temporal-Embodiment Memory (STEM), integrating spatial scene geometry, temporal event history, and embodiment profiles into a shared representation. High-level brain model performs global planning by retrieving and updating STEM while low-level controllers execute actions locally. Enables dynamic task allocation, fault-tolerant collaboration, and consistent state synchronization. Tested across restaurants, supermarkets, and households with heterogeneous embodiments.