Incremental Few-Shot Adaptation for Non-Prehensile Object Manipulation using Parallelizable Physics Simulators

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Baumeister, Fabian, Mack, Lukas, Stueckler, Joerg
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909556566654976
author Baumeister, Fabian
Mack, Lukas
Stueckler, Joerg
author_facet Baumeister, Fabian
Mack, Lukas
Stueckler, Joerg
contents Few-shot adaptation is an important capability for intelligent robots that perform tasks in open-world settings such as everyday environments or flexible production. In this paper, we propose a novel approach for non-prehensile manipulation which incrementally adapts a physics-based dynamics model for model-predictive control (MPC). The model prediction is aligned with a few examples of robot-object interactions collected with the MPC. This is achieved by using a parallelizable rigid-body physics simulation as dynamic world model and sampling-based optimization of the model parameters. In turn, the optimized dynamics model can be used for MPC using efficient sampling-based optimization. We evaluate our few-shot adaptation approach in object pushing experiments in simulation and with a real robot.
format Preprint
id arxiv_https___arxiv_org_abs_2409_13228
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Incremental Few-Shot Adaptation for Non-Prehensile Object Manipulation using Parallelizable Physics Simulators
Baumeister, Fabian
Mack, Lukas
Stueckler, Joerg
Robotics
Machine Learning
Few-shot adaptation is an important capability for intelligent robots that perform tasks in open-world settings such as everyday environments or flexible production. In this paper, we propose a novel approach for non-prehensile manipulation which incrementally adapts a physics-based dynamics model for model-predictive control (MPC). The model prediction is aligned with a few examples of robot-object interactions collected with the MPC. This is achieved by using a parallelizable rigid-body physics simulation as dynamic world model and sampling-based optimization of the model parameters. In turn, the optimized dynamics model can be used for MPC using efficient sampling-based optimization. We evaluate our few-shot adaptation approach in object pushing experiments in simulation and with a real robot.
title Incremental Few-Shot Adaptation for Non-Prehensile Object Manipulation using Parallelizable Physics Simulators
topic Robotics
Machine Learning
url https://arxiv.org/abs/2409.13228