Cascaded Gaussian Processes for Data-efficient Robot Dynamics Learning

TL;DR

This paper introduces a cascaded Gaussian process framework inspired by robot dynamics, significantly improving data efficiency and learning speed for robot inverse dynamics modeling, tested on multiple manipulators and motions.

Contribution

The paper proposes a novel cascaded Gaussian process approach for robot inverse dynamics, reducing dimensionality and enhancing data efficiency and generalization capabilities.

Findings

Inward cascaded GP outperforms standard GP in learning speed

Cascaded GP improves data efficiency in robot dynamics learning

Experimental validation on Jaco 2 and SARCOS manipulators shows consistent benefits

Abstract

Motivated by the recursive Newton-Euler formulation, we propose a novel cascaded Gaussian process learning framework for the inverse dynamics of robot manipulators. This approach leads to a significant dimensionality reduction which in turn results in better learning and data efficiency. We explore two formulations for the cascading: the inward and outward, both along the manipulator chain topology. The learned modeling is tested in conjunction with the classical inverse dynamics model (semi-parametric) and on its own (non-parametric) in the context of feed-forward control of the arm. Experimental results are obtained with Jaco 2 six-DOF and SARCOS seven-DOF manipulators for randomly defined sinusoidal motions of the joints in order to evaluate the performance of cascading against the standard GP learning. In addition, experiments are conducted using Jaco 2 on a task emulating a pouring maneuver. Results indicate a consistent improvement in learning speed with the inward cascaded GP model and an overall improvement in data efficiency and generalization.