Neural Robot Dynamics

TL;DR

This paper introduces NeRD, a neural simulator for articulated robots that predicts dynamics accurately, generalizes across tasks, and can be fine-tuned with real data, improving simulation and policy learning.

Contribution

NeRD replaces traditional dynamics solvers with a neural model that generalizes across tasks and environments, enabling stable, accurate, and adaptable robot simulations.

Findings

NeRD achieves stable and accurate predictions over 1000 simulation steps.

NeRD generalizes across different tasks and environments.

NeRD can be fine-tuned with real-world data to improve simulation fidelity.

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.