Localization in a wide range of industrial environments using relative 3D ceiling features

This paper presents a localization system for mobile robots that are able to navigate autonomously in industrial environments like factory and exhibition halls. Previous approaches show that ceiling structures like beams, pipes and lighting installation are well-suited for self-localization in large halls. With this paper we describe how to use ceiling structures, measured with a 3D laser range sensor, as natural landmarks. Due to the novel relative representation of 3D ceiling features this method can be used in a wide range of industrial halls with sloped, stepped and even curved ceiling structures. In addition to this new algorithm the paper presents experimental results with an autonomous fork lift working without the need for artificial landmarks or extra infrastructure.