Conference on Robot Learning (CoRL) 2025

Neural Robot Dynamics

NVIDIA
We demonstrate our approach by training NeRD models on six diverse robotic
                 systems. We show five of them here from left: Cartpole, Double Pendulum, Ant, Franka, ANYmal.

Abstract

Accurate and efficient simulation of modern robots remains challenging due to their high degrees of freedom and intricate mechanisms. Neural simulators have emerged as a promising alternative to traditional analytical simulators, capable of efficiently predicting complex dynamics and adapting to real-world data; however, existing neural simulators typically require application-specific training and fail to generalize to novel tasks and/or environments, primarily due to inadequate representations of the global state. In this work, we address the problem of learning generalizable neural simulators for robots that are structured as articulated rigid bodies. We propose NeRD (Neural Robot Dynamics), learned robot-specific dynamics models for predicting future states for articulated rigid bodies under contact constraints. NeRD uniquely replaces the low-level dynamics and contact solvers in an analytical simulator and employs a robot-centric and spatially-invariant simulation state representation. We integrate the learned NeRD models as an interchangeable backend solver within a state-of-the-art robotics simulator. We conduct extensive experiments to show that the NeRD simulators are stable and accurate over a thousand simulation steps; generalize across tasks and environment configurations; enable policy learning exclusively in a neural engine; and, unlike most classical simulators, can be fine-tuned from real-world data to bridge the gap between simulation and reality.

Overview

NeRD is characterized by two key innovations: (1) hybrid prediction framework (subfigures (a)(b)), where a neural network model uniquely replaces only the application-agnostic simulation modules in a conventional simulator, i.e., the low-level forward dynamics and contact solvers (subfigures (a)(b)), for the generalization across tasks and environments; (2) robot-centric state representation (subfigure (c)), to ensure generalizability on robot’s spatial locations.


NeRD adopts a hybrid prediction framework. NeRD replaces the physics solver in an analytical simulator, and takes intermediate simulation quantities as inputs including robot state, contact information, and joint-space torques. The inputs are all represented in the robot base frame.

Framework overview for Neural Robot Dynamics (NeRD). (a) Workflow of a classical robotics simulator. The quantities shaded in green are application-agnostic. (b) Hybrid prediction framework of the NeRD-integrated simulator. Inputs to NeRD are the robot-centric state representations (illustrated in (c)) within a history window.

More Visual Results

Double Pendulum with Various Grounds

A single NeRD model for double pendulum genearlizes across various contact configurations

(blue rollouts are simulated by ground-truth simulator, orange rollouts are simulated by NeRD)




Policy Learning and Evaluation

Robot policies can be learned exclusively in the NeRD-integrated neural simulator.

The executions of those learned policies exhibit high matching between the NeRD-integrated simulator and the ground-truth analytical simulator.

(Each robot instance shares a single NeRD model)




Franka Reach Policy in Real

The Franka reach policy learned in the NeRD-integrated neural simulator can be zero-shot deployed in real




Fine-tuning from Real Data

The NeRD models pre-trained from simulators can be fine-tuned from real data to capture real-world dynamics.


We fine-tuned a NeRD model (pre-trained from analytical simulation data) for cube tossing dynamics using real-world data. The simulator integrated with the fine-tuned NeRD model better matches real-world cube tossing dynamics.

BibTeX

@article{xu2025nerd,
  author    = {Xu, Jie and Heiden, Eric and Akinola, Iretiayo and Fox, Dieter and Macklin, Miles and Narang, Yashraj},
  title     = {Neural Robot Dynamics},
  journal   = {CoRL},
  year      = {2025},
}