A basic algorithm and data structures for sensor-based path planning in unknown environments

Abstract

Deals with the sensor-based path planning problem. In particular, we present the case for manipulator arms. We assume that the knowledge of workspace is partially known and that it can be increased by using a laser-like sensor. We propose a representation of the workspace that is well adapted to exploit the information obtained from the sensors and then compute collision detection. A sensor-based version of the Ariadne's Clew Algorithm is used to incrementally search for the free space and compute a path to a goal configuration.