3D LiDAR and 1D Radar Distance Determination for Safe Navigation of Automated Vessels in Lock Basins

Abstract

Highly automated vessel operations require sensor systems for environmental perception to ensure safe navigation, especially in confined areas like lock basins. This paper evaluates distance estimations using two different sensor systems, namely 1D Frequency-Modulated Continuous-Wave (FMCW) Radars and a 3D LiDAR, when entering a lock basin with a test vessel. A main contribution of this study is to quantitatively evaluate the accuracy of distances derived from data of both sensor types. To do so, error metrics are introduced and calculated. Additionally, a fusion algorithm is applied to fuse multiple Radar sensors as well as combine LiDAR and Radar data to improve distance estimation. For evaluating the distance accuracy, the nominal lock width serves as ground truth. Error metrics show that distance accuracy is increased by fusing Radar and LiDAR data. The RMSE is less than or equal to 0.13 m for all evaluated measurements. This confirms the benefits of Radar and LiDAR sensors in automated vessel navigation, showing the potential of these technologies for collision avoidance with critical infrastructure like locks.