Common ROS SLAM Packages (GMapping / SLAM Toolbox / RTAB-Map)

Three ready-to-run ROS packages for mapping and localization that just need a config file, not custom code.

GMapping is the oldest 2D lidar SLAM package (particle-filter-based, originating from OpenSLAM), and was the default choice for teaching robots in the ROS 1 era. SLAM Toolbox, developed by Steve Macenski, is the main option for 2D lidar mapping in ROS 2 today; it's graph-optimization-based and supports continuing to map on top of an existing map, as well as large environments. RTAB-Map, developed by Mathieu Labbé, targets RGB-D and stereo cameras, with built-in appearance-based loop-closure detection, and can output a dense 3D map. Their value is in turning SLAM into something you can get running just by changing a configuration file, producing a navigable occupancy grid map that plugs directly into Nav2; the tradeoff is that accuracy and robustness fall short of an algorithm specially tuned for a given environment.

ExampleRun SLAM Toolbox on a mobile base with a 2D lidar to build a grid map of a building floor, then hand it to Nav2 for navigation.

Related

Simultaneous Localization and Mapping · LiDAR SLAM · Occupancy Grid Map · ROS 2 Navigation Stack (Nav2) · RTAB-Map · Package (ROS)

Sources

slam_toolbox - GitHub
gmapping - ROS Wiki
RTAB-Map