 |
International Society of Science and Applied Technologies |
| A Study on the Development of Route Recommendation Based on Trajectory Clustering for Autonomous Ship | ||||
| Author | Jungyeon Choi
|
|||
| Co-Author(s) | Taewoong Hwang; Hyoseon Hwang; Ik-Hyun Youn
|
|||
| Abstract | The development of autonomous ships promises to revolutionize maritime transportation by enhancing efficiency, safety, and reducing human error. A crucial aspect of autonomous navigation is the ability to recommend optimal routes that minimize fuel consumption, reduce travel time, and avoid hazards. This study explores a route recommendation system for autonomous ships based on trajectory clustering. Using K-means and DBSCAN clustering algorithms, distinct ship trajectory patterns were identified to improve route accuracy. DBSCAN effectively handled noisy, complex data and detected arbitrarily shaped clusters, while K-means excelled with well-defined, regular trajectories. Additionally, an automated altering point detection system was incorporated to adapt to changes in ship movement and adjust routes. This study highlights the potential of data-driven techniques to enhance the safety and efficiency of autonomous ship navigation. Future work should focus on improving data handling, especially for noisy or incomplete data, and incorporating real-time environmental inputs to further optimize route planning.
|
|||
| Keywords | Autonomous ship, Trajectory clustering, K-means, DBSCAN, MASS | |||
| Article #: RQD2025-143 | ||||
Proceedings of 30th ISSAT International Conference on Reliability & Quality in Design |