Work

Empirical Probabilistic Roadmap Informed by Automatic Identification System Trajectories for Ship Path Planning Under Tidal Variations in Coastal Waters

Year: 2026

Type: article

Abstract: Coastal waters are characterized by complex spatial constraints, including narrow channels, irregular shorelines, and shallow-water areas. To address these challenges, this study proposes an Empirical Probabilistic Roadmap (EPRM) method for ship path planning based on Automatic Identification System (AIS) trajectory data. Historical AIS data of a regularly operated coastal passenger ferry near Mokpo, Republic of Korea, were cleaned, interpolated, and resampled at common longitude intervals. An empirical navigable space was then generated from the outer boundaries of the trajectories, and nodes were preferentially sampled within it according to the trajectory density. In addition, predicted tidal levels were compared with an empirically derived tidal threshold to select an appropriate empirical navigable space for the time of operation, thereby reflecting historically observed tide-dependent route selection in the path-planning process. The proposed method was compared with four benchmark path-planning algorithms, namely the A* graph-search algorithm, the Probabilistic Roadmap (PRM), the Rapidly-exploring Random Tree (RRT), and its asymptotically optimal variant RRT*, in terms of path length, search time, and similarity to actual AIS trajectories. Trajectory similarity was assessed using Modified Hausdorff Distance (MHD), Dynamic Time Warping (DTW), and Longest Common Subsequence (LCSS). The results showed that the EPRM generated paths with small run-to-run variability and the highest similarity to actual AIS trajectories among the compared algorithms under the study conditions, although its computational advantage was limited to the path-query phase once the roadmap had been constructed. Moreover, different paths to the same destination were generated as the normal and detour empirical navigable spaces were selected according to tidal conditions. It should be noted that the AIS-derived space represents historically used waters rather than a verified safe-navigation area, and its integration with bathymetric and under-keel clearance information is required for operational use.

Source: Journal of Marine Science and Engineering

DOI: 10.3390/jmse14212114

Authors: Su-Bong Kwak, Da-Un Jang, Chan-Sub Lee

Institutions: Mokpo National Maritime University, Korea Institute of Maritime and Fisheries Technology

Language: English

Topic: Robotic Path Planning Algorithms

Subfield: Control and Systems Engineering

Field: Engineering

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: gold

Estimated APC paid: $3,137