Glossary
SLAM
Simultaneous Localization and Mapping — building a map of an unknown environment while keeping track of your position in it.
SLAM (Simultaneous Localization and Mapping) is the problem of building a map and figuring out where you are in that map at the same time. The classic application is a mobile robot in a new environment.
Two main flavors based on the sensor:
- Visual SLAM (vSLAM) uses cameras. ORB-SLAM3 is the de facto open-source baseline.
- Lidar SLAM uses laser scanners. LOAM and its descendants dominate.
Newer approaches use NeRFs or Gaussian splatting to build photorealistic 3D maps. They're better for offline mapping than online SLAM as of 2026.
For mobile-base navigation in ROS 2, Nav2 + slam_toolbox is the standard stack.