Computation of Collision Distance and Gradient using an Automatic Sphere Approximation of the Robot Model with Bounded Error

Conference: ISR 2018 - 50th International Symposium on Robotics
06/20/2018 - 06/21/2016 at München, Germany

Proceedings: ISR 2018

Pages: 8Language: englishTyp: PDF

Personal VDE Members are entitled to a 10% discount on this title

Voelz, Andreas; Graichen, Knut (Institute of Measurement, Control and Microtechnology, Ulm University, Ulm, Germany)

This paper presents a method for computing the distance to collision and its gradient, which is a time-critical part of optimization-based motion planners like CHOMP. The computation is based on a sphere approximation of the robot model and an Euclidean distance transformation of the voxelized environment. As speed and accuracy depend on the number of spheres, an algorithm is proposed for error-bounded approximation of a simplified robot model.