ECHO: Edge-Cloud Humanoid Orchestration for Language-to-Motion Control

2026-03-17

AI Summary

This work proposes an efficient, safe, and real-time method for translating natural language instructions into full-body motions for humanoid robots. Leveraging an edge-cloud collaborative architecture, the cloud generates language-conditioned motion sequences using a diffusion model (1D UNet + CLIP + DDIM), while the edge executes them in closed-loop via a reinforcement learning policy. The approach introduces a novel 38-dimensional native robot motion representation that eliminates runtime retargeting. Efficient sim-to-real transfer is achieved through teacher-student distillation, evidence-aware adaptation, and morphological symmetry constraints. Integrated IMU-based fall detection and autonomous recovery further enhance robustness. Evaluated on HumanML3D, the system achieves an FID of 0.029 and R-Precision Top-1 of 0.686, and demonstrates stable, diverse instruction-following on the Unitree G1 robot without any hardware-specific fine-tuning.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Natural Language Processing: Safety and Robustness
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

We present ECHO, an edge--cloud framework for language-driven whole-body control of humanoid robots. A cloud-hosted diffusion-based text-to-motion generator synthesizes motion references from natural language instructions, while an edge-deployed reinforcement-learning tracker executes them in closed loop on the robot. The two modules are bridged by a compact, robot-native 38-dimensional motion representation that encodes joint angles, root planar velocity, root height, and a continuous 6D root orientation per frame, eliminating inference-time retargeting from human body models and remaining directly compatible with low-level PD control. The generator adopts a 1D convolutional UNet with cross-attention conditioned on CLIP-encoded text features; at inference, DDIM sampling with 10 denoising steps and classifier-free guidance produces motion sequences in approximately one second on a cloud GPU. The tracker follows a Teacher--Student paradigm: a privileged teacher policy is distilled into a lightweight student equipped with an evidential adaptation module for sim-to-real transfer, further strengthened by morphological symmetry constraints and domain randomization. An autonomous fall recovery mechanism detects falls via onboard IMU readings and retrieves recovery trajectories from a pre-built motion library. We evaluate ECHO on a retargeted HumanML3D benchmark, where it achieves strong generation quality (FID 0.029, R-Precision Top-1 0.686) under a unified robot-domain evaluator, while maintaining high motion safety and trajectory consistency. Real-world experiments on a Unitree G1 humanoid demonstrate stable execution of diverse text commands with zero hardware fine-tuning.

Problem

• language-to-motion
• humanoid control
• edge-cloud orchestration
• zero-shot execution
• motion safety

Innovation

• edge-cloud orchestration
• diffusion-based text-to-motion
• robot-native motion representation
• sim-to-real transfer
• autonomous fall recovery