Modelling and Learning Dynamics for Robotic Food-Cutting
(2021) p.1194-1200- Abstract
- Interaction dynamics are difficult to model analytically, making data-driven controllers preferable for contact-rich manipulation tasks. In this work, we approximate the intricate dynamics of food-cutting with a Long Short-Term Memory (LSTM) model to apply a Model Predictive Controller (MPC). We propose a problem formulation that allows velocity-controlled robots to learn the interaction dynamics and tackle the difficulty of multi-step predictions by training the model with a horizon curriculum. We experimentally demonstrate that our approach leads to good predictive performance that scales for longer prediction horizons, generalizes to unseen object classes and results in controller behaviors with an understanding of the cutting dynamics.
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/f3f8a772-e6d5-4dec-b4e0-60b2ea2f7065
- author
- Mitsioni, Ioanna
; Karayiannidis, Yiannis
LU
and Kragic, Danica
- publishing date
- 2021
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- 2021 IEEE 17th International Conference on Automation Science and Engineering (CASE)
- pages
- 7 pages
- external identifiers
-
- scopus:85117040503
- DOI
- 10.1109/CASE49439.2021.9551558
- language
- English
- LU publication?
- no
- id
- f3f8a772-e6d5-4dec-b4e0-60b2ea2f7065
- date added to LUP
- 2022-12-14 15:07:36
- date last changed
- 2025-04-04 14:44:01
@inproceedings{f3f8a772-e6d5-4dec-b4e0-60b2ea2f7065, abstract = {{Interaction dynamics are difficult to model analytically, making data-driven controllers preferable for contact-rich manipulation tasks. In this work, we approximate the intricate dynamics of food-cutting with a Long Short-Term Memory (LSTM) model to apply a Model Predictive Controller (MPC). We propose a problem formulation that allows velocity-controlled robots to learn the interaction dynamics and tackle the difficulty of multi-step predictions by training the model with a horizon curriculum. We experimentally demonstrate that our approach leads to good predictive performance that scales for longer prediction horizons, generalizes to unseen object classes and results in controller behaviors with an understanding of the cutting dynamics.}}, author = {{Mitsioni, Ioanna and Karayiannidis, Yiannis and Kragic, Danica}}, booktitle = {{2021 IEEE 17th International Conference on Automation Science and Engineering (CASE)}}, language = {{eng}}, pages = {{1194--1200}}, title = {{Modelling and Learning Dynamics for Robotic Food-Cutting}}, url = {{http://dx.doi.org/10.1109/CASE49439.2021.9551558}}, doi = {{10.1109/CASE49439.2021.9551558}}, year = {{2021}}, }