RLinf-USER: A Unified and Extensible System for Real-World Online Policy Learning in Embodied AI

TL;DR

USER is a comprehensive system that facilitates real-world online policy learning for embodied AI by integrating hardware management, adaptive communication, and flexible learning frameworks, enabling scalable, long-horizon experiments across diverse robots and models.

Contribution

The paper introduces USER, a novel unified system that addresses hardware heterogeneity, communication challenges, and long-term training needs for real-world online policy learning in embodied AI.

Findings

Enables multi-robot coordination in real-world settings.

Supports heterogeneous manipulators and large vision-language models.

Facilitates long-horizon asynchronous training with robust crash recovery.

Abstract

Online policy learning directly in the physical world is a promising yet challenging direction for embodied intelligence. Unlike simulation, real-world systems cannot be arbitrarily accelerated, cheaply reset, or massively replicated, which makes scalable data collection, heterogeneous deployment, and long-horizon effective training difficult. These challenges suggest that real-world policy learning is not only an algorithmic issue but fundamentally a systems problem. We present USER, a Unified and extensible SystEm for Real-world online policy learning. USER treats physical robots as first-class hardware resources alongside GPUs through a unified hardware abstraction layer, enabling automatic discovery, management, and scheduling of heterogeneous robots. To address cloud-edge communication, USER introduces an adaptive communication plane with tunneling-based networking, distributed data channels for traffic localization, and streaming-multiprocessor-aware weight synchronization to regulate GPU-side overhead. On top of this infrastructure, USER organizes learning as a fully asynchronous framework with a persistent, cache-aware buffer, enabling efficient long-horizon experiments with robust crash recovery and reuse of historical data. In addition, USER provides extensible abstractions for rewards, algorithms, and policies, supporting online imitation or reinforcement learning of CNN/MLP, generative policies, and large vision-language-action (VLA) models within a unified pipeline. Results in both simulation and the real world show that USER enables multi-robot coordination, heterogeneous manipulators, edge-cloud collaboration with large models, and long-running asynchronous training, offering a unified and extensible systems foundation for real-world online policy learning.