Toward Hardware-Agnostic Quadrupedal World Models via Morphology Conditioning

2026-04-09

AI Summary

This work addresses the limited cross-platform generalization of existing quadrupedal world models, which are overly dependent on specific hardware morphologies and require retraining upon body changes. To overcome this, the authors propose a morphology-conditioned Quadrupedal World Model (QWM) that explicitly encodes physical robot parameters to decouple environmental dynamics from body morphology. This approach enables, for the first time, zero-shot control of quadrupedal world models on unseen morphologies, circumventing the adaptation delays inherent in traditional implicit system identification. By integrating a morphology encoder, a reward normalization module, and a generative dynamics model, QWM functions as a universal neural simulator across a distribution of quadrupedal morphologies, substantially enhancing model generalization and deployment efficiency.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Intelligent Robots: Embodied AI
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

World models promise a paradigm shift in robotics, where an agent learns the underlying physics of its environment once to enable efficient planning and behavior learning. However, current world models are often hardware-locked specialists: a model trained on a Boston Dynamics Spot robot fails catastrophically on a Unitree Go1 due to the mismatch in kinematic and dynamic properties, as the model overfits to specific embodiment constraints rather than capturing the universal locomotion dynamics. Consequently, a slight change in actuator dynamics or limb length necessitates training a new model from scratch. In this work, we take a step towards a framework for training a generalizable Quadrupedal World Model (QWM) that disentangles environmental dynamics from robot morphology. We address the limitations of implicit system identification, where treating static physical properties (like mass or limb length) as latent variables to be inferred from motion history creates an adaptation lag that can compromise zero-shot safety and efficiency. Instead, we explicitly condition the generative dynamics on the robot's engineering specifications. By integrating a physical morphology encoder and a reward normalizer, we enable the model to serve as a neural simulator capable of generalizing across morphologies. This capability unlocks zero-shot control across a range of embodiments. We introduce, for the first time, a world model that enables zero-shot generalization to new morphologies for locomotion. While we carefully study the limitations of our method, QWM operates as a distribution-bounded interpolator within the quadrupedal morphology family rather than a universal physics engine, this work represents a significant step toward morphology-conditioned world models for legged locomotion.

Problem

• world models
• quadrupedal locomotion
• morphology generalization
• hardware-agnostic
• embodiment mismatch

Innovation

• world model
• morphology conditioning
• zero-shot generalization
• quadrupedal locomotion
• hardware-agnostic