Analysis of 3D LiDAR and 1D FMCW Radar Effectiveness for Distance Estimation in Inland Ports for Remote-Controlled Ship Navigation

Kurzfassung

Vessel safety during remote-controlled ship navigation (RCSN) is compromised by depth ambiguity when relying solely on 2D camera streams. While radar and LiDAR sensors have been proposed to increase situation awareness, practical performance comparisons for RCSN applications are sparse. This study proposes a practical evaluation of 1D FMCW radar and near-field LiDAR sensors for RCSN, focusing on their effectiveness in distance estimation to static objects within harbor settings. A novel LiDAR processing technique is developed, integrating water-plane filtering and surface normal estimation for robust distance estimation. For evaluation, we conduct sea trials by mounting both sensors on a vessel and systematically compare sensor performance during a prototypic maneuver. Results indicate that radar is less likely distracted by elusive objects, while LiDAR achieves high-resolution sampling with increased accuracy detecting even weak reflecting objects. These findings highlight specific use cases for each sensor in RCSN and provide additional guidance for their optimal deployment to enhance operator situation awareness.