Simultaneous Localization and Mapping

A foundational robotics technique in which a robot constructs a map of an unknown environment while simultaneously estimating its own pose within that map. SLAM combines sensor input (typically LiDAR, depth cameras, or monocular vision) with probabilistic state estimation (Kalman filters, particle filters, factor graphs) and loop-closure detection to produce a consistent map and localised trajectory. In accessible-navigation systems for blind users, SLAM enables map-less guide robots to operate in venues where no pre-built map exists, since the robot can build the map online during the walk. Common open-source implementations include Cartographer ROS, ORB-SLAM, RTAB-Map, and Hector SLAM.