Zero-Shot Parameter Learning of Robot Dynamics Using Bayesian Statistics and Prior Knowledge

TL;DR

This paper introduces a Bayesian-based zero-shot learning method for robot dynamics parameter identification that leverages prior knowledge to reduce measurement requirements and improve generalization.

Contribution

It presents a novel Bayesian approach that incorporates prior information for robot parameter learning, enabling accurate identification with minimal or no additional measurements.

Findings

Successfully learned inertial, mechanical, and base parameters of a robot

Ensured physical feasibility and confidence intervals in estimates

Provided priors for serial robots without datasheets or CAD models

Abstract

Inertial parameter identification of industrial robots is an established process, but standard methods using Least Squares or Machine Learning do not consider prior information about the robot and require extensive measurements. Inspired by Bayesian statistics, this paper presents an identification method with improved generalization that incorporates prior knowledge and is able to learn with only a few or without additional measurements (Zero-Shot Learning). Furthermore, our method is able to correctly learn not only the inertial but also the mechanical and base parameters of the MABI Max 100 robot while ensuring physical feasibility and specifying the confidence intervals of the results. We also provide different types of priors for serial robots with 6 degrees of freedom, where datasheets or CAD models are not available.