| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
The gmapping package implements an easy-to-use SLAM algorithm, but it seems to assume a static environment. If the room changes for any reason, it easily confuses the algorithm. How do you allow a robot to maintain its map but also handle events like, a door being closed or a person walking infront of the laser scanner, without gmapping getting lost or destroying its map?
Actually, gmapping is quite stable, when configured correctly, with respect to changes in the environment.
Typically, you only map once and the use localization on the created map using e.g. amcl which is even more agnostic to changes.
But you are right, if the environment changes much during mapping, these are just not the right algorithms. You would need to use other algorithms providing long-term mapping capabilities. But AFAIK there are currently non available in ROS.
Asked: 2016-08-17 13:42:10 -0500
Seen: 87 times
Last updated: Aug 18 '16
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
