A Data-Efficient Approach to Precise and Controlled Pushing

Maria Bauza, Francois R. Hogan, Alberto Rodriguez
; Proceedings of The 2nd Conference on Robot Learning, PMLR 87:336-345, 2018.

Abstract

Decades of research in control theory have shown that simple controllers, when provided with timely feedback, can control complex systems. Pushing is an example of a complex mechanical system that is difficult to model accurately due to unknown system parameters such as coefficients of friction and pressure distributions. In this paper, we explore the data-complexity required for controlling, rather than modeling, such a system. Results show that a model-based control approach, where the dynamical model is learned from data, is capable of performing complex pushing trajectories with a minimal amount of training data (<10 data points). The dynamics of pushing interactions are modeled using a Gaussian process (GP) and are leveraged within a model predictive control approach that linearizes the GP and imposes actuator and task constraints for a planar manipulation task.

Cite this Paper


BibTeX
@InProceedings{pmlr-v87-bauza18a, title = {A Data-Efficient Approach to Precise and Controlled Pushing}, author = {Bauza, Maria and Hogan, Francois R. and Rodriguez, Alberto}, booktitle = {Proceedings of The 2nd Conference on Robot Learning}, pages = {336--345}, year = {2018}, editor = {Aude Billard and Anca Dragan and Jan Peters and Jun Morimoto}, volume = {87}, series = {Proceedings of Machine Learning Research}, address = {}, month = {29--31 Oct}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v87/bauza18a/bauza18a.pdf}, url = {http://proceedings.mlr.press/v87/bauza18a.html}, abstract = {Decades of research in control theory have shown that simple controllers, when provided with timely feedback, can control complex systems. Pushing is an example of a complex mechanical system that is difficult to model accurately due to unknown system parameters such as coefficients of friction and pressure distributions. In this paper, we explore the data-complexity required for controlling, rather than modeling, such a system. Results show that a model-based control approach, where the dynamical model is learned from data, is capable of performing complex pushing trajectories with a minimal amount of training data (<10 data points). The dynamics of pushing interactions are modeled using a Gaussian process (GP) and are leveraged within a model predictive control approach that linearizes the GP and imposes actuator and task constraints for a planar manipulation task. } }
Endnote
%0 Conference Paper %T A Data-Efficient Approach to Precise and Controlled Pushing %A Maria Bauza %A Francois R. Hogan %A Alberto Rodriguez %B Proceedings of The 2nd Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2018 %E Aude Billard %E Anca Dragan %E Jan Peters %E Jun Morimoto %F pmlr-v87-bauza18a %I PMLR %J Proceedings of Machine Learning Research %P 336--345 %U http://proceedings.mlr.press %V 87 %W PMLR %X Decades of research in control theory have shown that simple controllers, when provided with timely feedback, can control complex systems. Pushing is an example of a complex mechanical system that is difficult to model accurately due to unknown system parameters such as coefficients of friction and pressure distributions. In this paper, we explore the data-complexity required for controlling, rather than modeling, such a system. Results show that a model-based control approach, where the dynamical model is learned from data, is capable of performing complex pushing trajectories with a minimal amount of training data (<10 data points). The dynamics of pushing interactions are modeled using a Gaussian process (GP) and are leveraged within a model predictive control approach that linearizes the GP and imposes actuator and task constraints for a planar manipulation task.
APA
Bauza, M., Hogan, F.R. & Rodriguez, A.. (2018). A Data-Efficient Approach to Precise and Controlled Pushing. Proceedings of The 2nd Conference on Robot Learning, in PMLR 87:336-345

Related Material