Navigation with Simultaneous Localization and Mapping: For Indoor Mobile Robot

Abstract

This thesis describes the development of an integrated solution for simultaneous localization and mapping (SLAM), path planning and path following for a four-wheel indoor robot. Running on the robot's on-board computer, the solution estimates maps and location of the robot in real-time.The maps built are occupancy grid maps, which are suitable for path planning and for human inspection. The implemented path planning algorithm is successful in generating paths which avoid obstacles and can be followed accurately by automatic control using a line-of-sight approach.Experiments are presented which demonstrate the software's ability to accurately build a map of many environments with no prior assumptions or preparations. Providing foundations for localization and navigation, the system represents both a proof of concept and a platform on which others can build further.