Understanding the Impact of Modeling Abstractions on Motion Planning for Deformable Linear Objects

Abstract

Robotic manipulation of deformable objects remains challenging due to the high dimensional configuration space and complex dynamics. In this work we demonstrate how the abstraction level used for modeling deformable objects can significantly impact the difficulty of the motion planning problem. We specifically focus on buckling—a nonlinear instability phenomenon that arises in response to compression of slender deformable objects. Using deformable linear objects (DLOs) as a case of study, we show that eliminating resistance to compression in the simulation model while penalizing compressed states in the planning objective increases both robustness and performance. We demonstrate our approach on a set of simulation examples and validate our results through physical robot experiments.