Revising Environment Maps in Real Time
No Thumbnail Available
Date
2022-04-14
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
When mapping indoor environments that have frequent changes in the environment, like whenever the state of doors is changed or if furniture has moved, existing robotic systems struggle since they most commonly build a static map of their surroundings that is not able to recognize its location whenever only a few things are changed in the surroundings. As robots continue to be used alongside humans, robots and robotic systems should be able to understand where in their environment they are without nearly as much previous knowledge and update their understanding of their surroundings as they explore. This research attempts to solve these problems and make a robot be able to localize itself in a previously known map and then dynamically update that map without the need for remapping the environment completely, providing itself the most up to date environment map for navigation purposes. This research yields a system that allows for the accurate revision of a previously generated ground truth environment map while being able to localize a robot with no previous knowledge of its location within the environment. This research also yields an implementation of this system on a physical robot with a LIDAR sensor, as well as a simulation of this system and a comparison of this method with other methods of localization and map revision of small, day-to-day changes to human indoor environments.
Description
Keywords
indoor maps, SLAM, simultaneous localization and mapping, LIDAR, mapping, robotics