In-Hand Object Rotation via Rapid Motor Adaptation

Haozhi Qi, Ashish Kumar, Roberto Calandra, Yi Ma, Jitendra Malik
Proceedings of The 6th Conference on Robot Learning, PMLR 205:1722-1732, 2023.

Abstract

Generalized in-hand manipulation has long been an unsolved challenge of robotics. As a small step towards this grand goal, we demonstrate how to design and learn a simple adaptive controller to achieve in-hand object rotation using only fingertips. The controller is trained entirely in simulation on only cylindrical objects, which then – without any fine-tuning – can be directly deployed to a real robot hand to rotate dozens of objects with diverse sizes, shapes, and weights over the z-axis. This is achieved via rapid online adaptation of the robot’s controller to the object properties using only proprioception history. Furthermore, natural and stable finger gaits automatically emerge from training the control policy via reinforcement learning. Code and more videos are available at https://github.com/HaozhiQi/Hora .

Cite this Paper

BibTeX
@InProceedings{pmlr-v205-qi23a, title = {In-Hand Object Rotation via Rapid Motor Adaptation}, author = {Qi, Haozhi and Kumar, Ashish and Calandra, Roberto and Ma, Yi and Malik, Jitendra}, booktitle = {Proceedings of The 6th Conference on Robot Learning}, pages = {1722--1732}, year = {2023}, editor = {Liu, Karen and Kulic, Dana and Ichnowski, Jeff}, volume = {205}, series = {Proceedings of Machine Learning Research}, month = {14--18 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v205/qi23a/qi23a.pdf}, url = {https://proceedings.mlr.press/v205/qi23a.html}, abstract = {Generalized in-hand manipulation has long been an unsolved challenge of robotics. As a small step towards this grand goal, we demonstrate how to design and learn a simple adaptive controller to achieve in-hand object rotation using only fingertips. The controller is trained entirely in simulation on only cylindrical objects, which then – without any fine-tuning – can be directly deployed to a real robot hand to rotate dozens of objects with diverse sizes, shapes, and weights over the z-axis. This is achieved via rapid online adaptation of the robot’s controller to the object properties using only proprioception history. Furthermore, natural and stable finger gaits automatically emerge from training the control policy via reinforcement learning. Code and more videos are available at https://github.com/HaozhiQi/Hora .} }
Endnote
%0 Conference Paper %T In-Hand Object Rotation via Rapid Motor Adaptation %A Haozhi Qi %A Ashish Kumar %A Roberto Calandra %A Yi Ma %A Jitendra Malik %B Proceedings of The 6th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2023 %E Karen Liu %E Dana Kulic %E Jeff Ichnowski %F pmlr-v205-qi23a %I PMLR %P 1722--1732 %U https://proceedings.mlr.press/v205/qi23a.html %V 205 %X Generalized in-hand manipulation has long been an unsolved challenge of robotics. As a small step towards this grand goal, we demonstrate how to design and learn a simple adaptive controller to achieve in-hand object rotation using only fingertips. The controller is trained entirely in simulation on only cylindrical objects, which then – without any fine-tuning – can be directly deployed to a real robot hand to rotate dozens of objects with diverse sizes, shapes, and weights over the z-axis. This is achieved via rapid online adaptation of the robot’s controller to the object properties using only proprioception history. Furthermore, natural and stable finger gaits automatically emerge from training the control policy via reinforcement learning. Code and more videos are available at https://github.com/HaozhiQi/Hora .
APA
Qi, H., Kumar, A., Calandra, R., Ma, Y. & Malik, J.. (2023). In-Hand Object Rotation via Rapid Motor Adaptation. Proceedings of The 6th Conference on Robot Learning, in Proceedings of Machine Learning Research 205:1722-1732 Available from https://proceedings.mlr.press/v205/qi23a.html.

Related Material