Surgical Robot Transformer (SRT): Imitation Learning for Surgical Tasks

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kim, Ji Woong, Zhao, Tony Z., Schmidgall, Samuel, Deguet, Anton, Kobilarov, Marin, Finn, Chelsea, Krieger, Axel
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917725516857344
author Kim, Ji Woong
Zhao, Tony Z.
Schmidgall, Samuel
Deguet, Anton
Kobilarov, Marin
Finn, Chelsea
Krieger, Axel
author_facet Kim, Ji Woong
Zhao, Tony Z.
Schmidgall, Samuel
Deguet, Anton
Kobilarov, Marin
Finn, Chelsea
Krieger, Axel
contents We explore whether surgical manipulation tasks can be learned on the da Vinci robot via imitation learning. However, the da Vinci system presents unique challenges which hinder straight-forward implementation of imitation learning. Notably, its forward kinematics is inconsistent due to imprecise joint measurements, and naively training a policy using such approximate kinematics data often leads to task failure. To overcome this limitation, we introduce a relative action formulation which enables successful policy training and deployment using its approximate kinematics data. A promising outcome of this approach is that the large repository of clinical data, which contains approximate kinematics, may be directly utilized for robot learning without further corrections. We demonstrate our findings through successful execution of three fundamental surgical tasks, including tissue manipulation, needle handling, and knot-tying.
format Preprint
id arxiv_https___arxiv_org_abs_2407_12998
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Surgical Robot Transformer (SRT): Imitation Learning for Surgical Tasks
Kim, Ji Woong
Zhao, Tony Z.
Schmidgall, Samuel
Deguet, Anton
Kobilarov, Marin
Finn, Chelsea
Krieger, Axel
Robotics
We explore whether surgical manipulation tasks can be learned on the da Vinci robot via imitation learning. However, the da Vinci system presents unique challenges which hinder straight-forward implementation of imitation learning. Notably, its forward kinematics is inconsistent due to imprecise joint measurements, and naively training a policy using such approximate kinematics data often leads to task failure. To overcome this limitation, we introduce a relative action formulation which enables successful policy training and deployment using its approximate kinematics data. A promising outcome of this approach is that the large repository of clinical data, which contains approximate kinematics, may be directly utilized for robot learning without further corrections. We demonstrate our findings through successful execution of three fundamental surgical tasks, including tissue manipulation, needle handling, and knot-tying.
title Surgical Robot Transformer (SRT): Imitation Learning for Surgical Tasks
topic Robotics
url https://arxiv.org/abs/2407.12998