2D LiDAR SLAM

Key Insight

Robots operating in unknown environments must construct a map while simultaneously tracking their position within it. This 2D LiDAR SLAM system pairs a front-end that matches scan points using algorithms like Iterative Closest Point (ICP) to estimate incremental motion with a back-end that maintains a pose graph. By detecting loop closures—recognizing when the robot has returned to a previously visited location—the system injects spatial constraints that correct accumulated drift, aligning the entire map and trajectory.