Multilevel Motion Planning

Project Goal

• Develop algorithms which can plan robot motions over multiple levels of abstraction while keeping formal guarantees like probabilistic completeness and asymptotic optimality.

Code

• An existing implementation of the quotient space planning framework is available in the Open Motion Planning Library (OMPL). Check out the overview and the tutorial.

Contributors

• • Andreas Orthey, Marc Toussaint, Sohaib Akbar, Adrien Escande, Eiichi Yoshida

Publications

• Andreas Orthey, Sohaib Akbar, Marc Toussaint, Multilevel Motion Planning: A Fiber Bundle Formulation, arXiv:2007.09435 [cs.RO], 2020, [PDF-preprint] [Bibtex]

• Andreas Orthey, Marc Toussaint, Section Patterns: Efficiently Solving Narrow Passage Problems in Multilevel Motion Planning, IEEE Transactions on Robotics, 2021, [PDF] [Bibtex]

• Andreas Orthey, Marc Toussaint, Sparse Multilevel Roadmaps for High-Dimensional Robotic Motion Planning, IEEE International Conference on Robotics and Automation (ICRA), 2021, [PDF] [BlogPost] [Bibtex]

• Andreas Orthey, Marc Toussaint, Rapidly-Exploring Quotient-Space Trees: Motion Planning using Sequential Simplifications, International Symposium on Robotics Research (ISRR), 2019, [PDF] [Bibtex]

• Andreas Orthey, Adrien Escande, Eiichi Yoshida, Quotient-Space Motion Planning, IEEE International Conference on Intelligent Robots and Systems (IROS), 2018, [PDF] [Bibtex]

#MultilevelMotionPlanning #QuotientSpacePlanning #SparseMultilevelRoadmaps