Algorithmic motion planning: The randomized approach

Abstract

Algorithms based on randomized sampling proved to be the only viable algorithmic tool for quickly solving motion planning problems involving many degrees of freedom. Information on the configuration space is acquired by generating samples and finding simple paths among them. Paths and samples are stored in a suitable data structure. According to this paradigm, in the recent years a wide number of algorithmic techniques have been proposed and some approaches are now widely used. This survey reviews the main algorithms, outlining their advantages and drawbacks, as well as the knowledge recently acquired in the field. © Springer-Verlag Berlin Heidelberg 2006.