Abstract:We are witnessing the rise of autonomous cars, which will likely revolutionize the way we travel. Arguably, the maritime domain lags behind, as ships operate on many more degrees of freedom (thus, a much larger search space): there is less physical infrastructure, and rules are less consistent and constraining than what is found on roads. The problem is further complicated by the inevitable co-existence of autonomous and human-operated ships: the latter may take unpredictable decisions, which require adjustments on the autonomous ones. Finally, the problem is inherently decentralised, there is no central authority, and communication means can be very diverse in terms of communication distance and performance, mandating special care on which information is shared and how. In this work, we elaborate on the challenges of trajectory prediction and adaptation for mixed autonomous and human-operated ships, and we propose initial ideas on potential approaches to address them.

What problem does this paper attempt to address?

This paper attempts to solve the problems of trajectory prediction and adaptation encountered by hybrid autonomous and manually - operated ships during maritime navigation. Specifically, the paper mainly focuses on the following aspects: 1. **Multi - degree - of - freedom operation**: Compared with land transportation, ship operations have more degrees of freedom (such as lateral, longitudinal, vertical directions, etc.), which leads to a larger search space and more complex navigation challenges. 2. **Co - existence problem**: Autonomous ships and human - operated ships will inevitably co - exist in the same water area. Human - operated ships may exhibit unpredictable behaviors, which pose higher adaptability requirements for autonomous ships. 3. **Decentralized system**: There is no central control agency, and communication means are diverse and have different performances (such as short - range Wi - Fi, medium - range 4G/5G networks, long - range LoRa radio communication, etc.). Therefore, how to select the most appropriate communication method and ensure effective information sharing is an important issue. 4. **Dynamically changing environment**: Factors including weather conditions, the actions of other ships, and unexpected events (such as accidents or floating objects) will all affect the navigation conditions and require real - time adjustment of trajectory planning. To solve these problems, the author proposes some preliminary ideas and technical approaches, mainly including: - **Field - based calculation method**: Use techniques such as gradient paths to achieve adaptive path planning. These paths can be quickly adjusted according to environmental changes to avoid collisions and respond in advance to changes in the intentions of other ships. - **Anisotropic field**: Dynamically adjust the content of data exchange according to different communication technologies and bandwidths. For example, high - definition video streams can be transmitted at close range, while only a small amount of key information is transmitted at long range. - **Error - perception field**: Introduce a reliability index to ensure that the cumulative error of data during transmission does not exceed the acceptable range, thereby improving the credibility of information. - **Runtime reconfiguration through fragmentation**: Decompose computational tasks into micro - activities so that they can be flexibly deployed according to the actually available resources. For example, some calculations can be completed on nearby better devices or even processed through the cloud. - **Learning - based method**: Use simulation training of collective behavior models and then execute them in a distributed environment. This method can be applied to multiple aspects such as trajectory calculation, ship behavior prediction, and selection of important data. In summary, this paper aims to explore how to implement an efficient autonomous ship navigation system in a complex and changeable marine environment, especially to ensure safe and reliable navigation in the case of the co - existence of manned and unmanned ships.

Towards adaptive trajectories for mixed autonomous and human-operated ships

New frontiers in the risk assessment of ship collision

An Outlook on the Future Marine Traffic Management System for Autonomous Ships

Hybrid Navigation Acceptability and Safety

COLREGs-Adaptive trajectory planning and decision-making in maritime autonomous surface ships

The human element in future Maritime Operations - perceived impact of autonomous shipping

Model Predictive Trajectory Optimization and Control for Autonomous Surface Vessels Considering Traffic Rules

Future of maritime autonomy: cybersecurity, trust and mariner's situational awareness

Enhanced Multimodal Trajectory Prediction for Autonomous Vehicles Using Advanced Diffusion Model Techniques

Human-autonomy collaboration in supervisory risk control of autonomous ships

Digital Twin of Autonomous Surface Vessels for Safe Maritime Navigation Enabled through Predictive Modeling and Reinforcement Learning

Towards supervisory risk control of autonomous ships

Vessel Trajectory Prediction in Maritime Transportation: Current Approaches and Beyond

Harnessing traditional controllers for fast-track training of deep reinforcement learning control strategies

AIS data-driven ship trajectory prediction modelling and analysis based on machine learning and deep learning methods

Regional ship behavior and trajectory prediction for maritime traffic management: A social generative adversarial network approach

Prediction-enabled path planning for multi-ship encounters in Oslofjord

A systematic review of human-AI interaction in autonomous ship systems

Multi-Ship Control and Collision Avoidance Using MPC and RBF-Based Trajectory Predictions

A Practical and Online Trajectory Planner for Autonomous Ships' Berthing, Incorporating Speed Control

Artificial Intelligence in Ship Trajectory Prediction