Neural Robot Dynamics

Jie Xu, Eric Heiden, Iretiayo Akinola, Dieter Fox, Miles Macklin, Yashraj Narang

Keywords: Robot Model Learning; Robotics Simulation; Neural Simulation

TL;DR: We learn a robot-centric neural physics module that is genearlizable across different tasks and environment setups for a particular robot.

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.

Submission Number: 222