Motion planning for the precise location of a mobile robot

Abstract

Motion planning for mobile vehicles in a sensor uncertainty environment is considered. The robot should visit a given set of targets while estimating its position by extended Kalman filtering based on distance measurements to known beacons. In a first part, piecewise linear paths are constructed that optimize the precision on the robot position when the targets are reached. The approach suggested simply relies on the introduction of additional targets in the definition of the robot mission. In a second part, the robot is allowed to mark out its route with beacons, in order to help estimate its own position. An approach is suggested that takes the uncertainty on the position of the dropped beacon into account. The dropping site is then optimized. Various examples are presented.