The field of 3D surface reconstruction and floorplan analysis is moving towards more robust and accurate methods for handling incomplete and noisy data. Recent developments have focused on leveraging edge-centric formulations, orthogonal projections, and semantic-geometric fusion to improve the reconstruction of floorplans and surfaces from point cloud data. Noteworthy papers include: CAGE, which proposes a continuity-aware edge network for robust floorplan reconstruction, achieving state-of-the-art performance on several datasets. FloorSAM, which integrates point cloud density maps with the Segment Anything Model for accurate floor plan reconstruction from LiDAR data, demonstrating better accuracy and robustness than traditional methods.