Collaborative SLAM in Dynamic Indoors: LiDAR Mapping with Fiducial Corrections

The growing use of mobile robots in dynamic indoor environments demands efficient and scalable Collaborative Simultaneous Localization and Mapping (C-SLAM) solutions. Traditional SLAM methods struggle with odometric drift in large-scale settings. This paper presents a distributed C-SLAM framework that combines LiDAR-based mapping, a drift-free Extended Kalman Filter, and a grid-based Iterative Closest Point algorithm. Fiducial Markers provide global pose corrections, while a hash-based sparse occupancy grid enables efficient map storage and sharing among robots. Experiments in both laboratory and industrial environments show a significant improvement of the localization errors compared to state-of-the-art techniques.