An on-the-fly evolutionary algorithm for robot motion planning

Computation of a collision-free path for a movable object among obstacles is an important problem in the fields of robotics. The simplest version of motion planning consists of generating a collision-free path for a movable object among known and static obstacles. In this paper, we introduce a two stage evolutionary algorithm. The first stage is designed to compute a collision-free path in a known environment. The second stage is designed to make on-the-fly updates of the robot current path according to the dynamic environmental modifications. Evolutionary techniques have proven to be useful to both quickly compute a new path and to take advantage of the initial path from the first stage. The tests have been made using simulations and a Lego Mindstorms Robot.