Can a robot navigate complex indoor spaces with limited sensor data? This paper presents 2DLIW-SLAM, a robust and accurate 2D LiDAR SLAM system designed for indoor mobile robots, overcoming motion degeneracy challenges in geometrically similar environments. This research investigates the robot's motion and navigation. The system processes LiDAR data through point and line extraction and establishes line-line constraints to augment sensor data. A tightly-coupled front-end integrates data from 2D LiDAR, an inertial measurement unit, and wheel odometry for real-time state estimation. The approach also uses a global feature point matching-based loop closure detection algorithm to mitigate front-end accumulated errors, building a globally consistent map. Experimental results show the system meets real-time requirements, exhibiting lower trajectory errors and greater robustness compared to existing methods. The open-source method promises enhanced navigation capabilities for indoor mobile robots, particularly in environments prone to degeneracy.
This work is appropriate for Measurement Science and Technology, a journal focused on advancements in measurement techniques and instrumentation. By presenting a multi-sensor fusion approach for 2D LiDAR SLAM, the paper fits the journal's scope. The system’s real-time performance and robustness contribute to the field of robotic navigation and mapping.