Learning How to Solve “Bubble Ball”

Hotae Lee, Monimoy Bujarbaruah, Francesco Borrelli

Proceedings of the 3rd Conference on Learning for Dynamics and Control, PMLR 144:1068-1079, 2021.

Abstract

“Bubble Ball” is a game built on a 2D physics engine, where a finite set of objects can modify the motion of a bubble-like ball. The objective is to choose the set and the initial configuration of the objects, in order to get the ball to reach a target flag. The presence of obstacles, friction, contact forces and combinatorial object choices make the game hard to solve. In this paper, we propose a hierarchical predictive framework which solves Bubble Ball. Geometric, kinematic and dynamic models are used at different levels of the hierarchy. At each level of the game, data collected during failed iterations are used to update models at all hierarchical level and converge to a feasible solution to the game. The proposed approach successfully solves a large set of Bubble Ball levels within reason-able number of trials. This proposed framework can also be used to solve other physics-based games, especially with limited training data from human demonstrations.

Cite this Paper

BibTeX


@InProceedings{pmlr-v144-lee21a,
  title = 	 {Learning How to Solve “Bubble Ball”},
  author =       {Lee, Hotae and Bujarbaruah, Monimoy and Borrelli, Francesco},
  booktitle = 	 {Proceedings of the 3rd Conference on Learning for Dynamics and Control},
  pages = 	 {1068--1079},
  year = 	 {2021},
  editor = 	 {Jadbabaie, Ali and Lygeros, John and Pappas, George J. and A. Parrilo, Pablo and Recht, Benjamin and Tomlin, Claire J. and Zeilinger, Melanie N.},
  volume = 	 {144},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {07 -- 08 June},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v144/lee21a/lee21a.pdf},
  url = 	 {https://proceedings.mlr.press/v144/lee21a.html},
  abstract = 	 {“Bubble Ball” is a game built on a 2D physics engine, where a finite set of objects can modify the motion of a bubble-like ball. The objective is to choose the set and the initial configuration of the objects, in order to get the ball to reach a target flag.  The presence of obstacles, friction, contact forces and combinatorial object choices make the game hard to solve. In this paper, we propose a hierarchical predictive framework which solves Bubble Ball. Geometric, kinematic and dynamic models are used at different levels of the hierarchy. At each level of the game, data collected during failed iterations are used to update models at all hierarchical level and converge to a feasible solution to the game. The proposed approach successfully solves a large set of Bubble Ball levels within reason-able number of trials. This proposed framework can also be used to solve other physics-based games, especially with limited training data from human demonstrations.}
}

Endnote

%0 Conference Paper
%T Learning How to Solve “Bubble Ball”
%A Hotae Lee
%A Monimoy Bujarbaruah
%A Francesco Borrelli
%B Proceedings of the 3rd Conference on Learning for Dynamics and Control
%C Proceedings of Machine Learning Research
%D 2021
%E Ali Jadbabaie
%E John Lygeros
%E George J. Pappas
%E Pablo A. Parrilo
%E Benjamin Recht
%E Claire J. Tomlin
%E Melanie N. Zeilinger	
%F pmlr-v144-lee21a
%I PMLR
%P 1068--1079
%U https://proceedings.mlr.press/v144/lee21a.html
%V 144
%X “Bubble Ball” is a game built on a 2D physics engine, where a finite set of objects can modify the motion of a bubble-like ball. The objective is to choose the set and the initial configuration of the objects, in order to get the ball to reach a target flag.  The presence of obstacles, friction, contact forces and combinatorial object choices make the game hard to solve. In this paper, we propose a hierarchical predictive framework which solves Bubble Ball. Geometric, kinematic and dynamic models are used at different levels of the hierarchy. At each level of the game, data collected during failed iterations are used to update models at all hierarchical level and converge to a feasible solution to the game. The proposed approach successfully solves a large set of Bubble Ball levels within reason-able number of trials. This proposed framework can also be used to solve other physics-based games, especially with limited training data from human demonstrations.

APA


Lee, H., Bujarbaruah, M. & Borrelli, F.. (2021). Learning How to Solve “Bubble Ball”. Proceedings of the 3rd Conference on Learning for Dynamics and Control, in Proceedings of Machine Learning Research 144:1068-1079 Available from https://proceedings.mlr.press/v144/lee21a.html.

Related Material

Download PDF