Load-Aware Locomotion Control for Humanoid Robots in Industrial Transportation Tasks

2026-03-15

AI Summary

This work addresses the challenges of dynamic coupling and partial observability in humanoid robots during industrial transport tasks, which arise from varying payloads and upper-body manipulation. To tackle these issues, the authors propose a decoupled yet coordinated mobile-manipulation control architecture that integrates kinematic reference trajectories, a history-based state estimator, and a highly conditioned joint-space residual policy. This framework encodes payload variations and disturbances into compact latent representations, enabling zero-shot sim-to-real transfer without fine-tuning. Experimental results demonstrate that the system achieves rapid training convergence, high-fidelity height tracking, and robust coordination between locomotion and manipulation, consistently across both simulation and physical platforms.

Technology Category

• Intelligent Robots: Manipulation
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Humanoid robots deployed in industrial environments are required to perform load-carrying transportation tasks that tightly couple locomotion and manipulation. However, achieving stable and robust locomotion under varying payloads and upper-body motions is challenging due to dynamic coupling and partial observability. This paper presents a load-aware locomotion framework for industrial humanoids based on a decoupled yet coordinated loco-manipulation architecture. Lower-body locomotion is controlled via a reinforcement learning policy producing residual joint actions on kinematically derived nominal configurations. A kinematics-based locomotion reference with a height-conditioned joint-space offset guides learning, while a history-based state estimator infers base linear velocity and height and encodes residual load- and manipulation-induced disturbances in a compact latent representation. The framework is trained entirely in simulation and deployed on a full-size humanoid robot without fine-tuning. Simulation and real-world experiments demonstrate faster training, accurate height tracking, and stable loco-manipulation. Project page: https://lequn-f.github.io/LALO/

Problem

• humanoid robots
• load-carrying
• locomotion control
• dynamic coupling
• partial observability

Innovation

• load-aware locomotion
• loco-manipulation
• reinforcement learning
• state estimation
• sim-to-real transfer