Toward Hardware-Agnostic Quadrupedal World Models via Morphology Conditioning
2026-04-09 • Robotics
RoboticsMachine Learning
AI summaryⓘ
The authors propose a new way to teach robots with four legs to move by making a model that understands both the environment and the robot's body shape. Instead of training a new model for each different robot, their method uses the robot's physical details directly to predict how it will move. This approach helps the model to quickly and safely control different types of four-legged robots without relearning from scratch. They show that their world model can generalize to new robot shapes but within a specific family of quadrupedal robots.
world modelsquadrupedal robotsrobot morphologygenerative dynamicszero-shot generalizationsystem identificationneural simulatorlocomotionrobot controlactuator dynamics
Authors
Mohamad H. Danesh, Chenhao Li, Amin Abyaneh, Anas Houssaini, Kirsty Ellis, Glen Berseth, Marco Hutter, Hsiu-Chin Lin
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.