Assume a certain task for a robot, for example: Move from point A to point B without colliding with any obstacle. Robot motion planning is the problem of finding a plan for the actuators of the robot such that the resulting movement of the robot satisfies the given task. The currently dominant class of algorithms for motion planning are sampling based, rapidly expanding a tree that explores the search space of possible plans. However, such algorithms usually use rough approximations the dynamics of the robotics, which may result in highly sub-optimal plans.
The goal of this topic is to design theory, algorithms, and software for efficiently handling realistic models of robot dynamics in sampling based motion planning. All algorithms will be implemented in the frame of the Open Motion Planning Library (OMPL) ) which will allow their usage by practitioners in robotics world-wide.