Local map design for local navigation planning

Abstract

This paper deals with the definition of robust tasks for displacement missions of a mobile robot in an indoor environment. For that, the concept of local map is introduced as a set of best landmarks used for planning and executing safe motions. The goal is to plan actions (path following) as well as observations (local maps), leading the robot to configurations where pertinent features can be sensed and thus allowing a best localization of the robot relative to its environment. At execution time, local navigation instead of global navigation is performed using local maps for the robot localization. For each mission, the definition of the local maps and their corresponding robot-tasks is realized automatically during the planning process. The local map design is based on a clustering method developed by Fisher (1958).