Safe Maneuvering, Efficient Navigation and Intelligent Management for Ships

1. Introduction

Maritime transport, serving as the cornerstone of global supply chains, facilitates over 80% of international trade by volume. However, this traditional industry is confronting multifaceted challenges: increasingly congested waterways due to growing maritime traffic elevate the risks of collisions and groundings; stringent environmental regulations from the International Maritime Organization (IMO) impose hard constraints on ship energy efficiency and carbon emissions [1]; and a global shortage of highly skilled seafarers is becoming increasingly apparent. In this context, research on safe ship maneuvering, efficient navigation, and intelligent management is no longer optional but a core driver propelling the shipping industry toward a safe, green, and efficient future. Cutting-edge technologies, exemplified by artificial intelligence (AI), big data analytics, and complex system modeling, are fundamentally reshaping the ship’s “perception–decision–action” loop [2]. This shift is transforming ship handling from an experience-dependent art into a science-based paradigm grounded in data and models.

2. An Overview of the Published Articles

Safety remains the top priority in shipping. While traditional safety measures heavily rely on human efforts, intelligent technologies aim to establish a proactive, data-driven safety framework based on human–machine collaboration. Current state-of-the-art research is advancing toward multi-modal information fusion. By integrating data from sources such as radar, computer vision, and AIS, modern intelligent ship systems can achieve all-weather perception beyond the visual horizon, enabling real-time detection and early warning of anomalous behaviors (e.g., deviation from fairways, dangerous approaches) for both their ship and target ships.

In this Special Issue, the study by Zhou et al. (Contribution 1) innovatively constructs an ontology model to identify abnormal ship behavior from a “navigation rule perspective.” This approach structurally integrates human nautical knowledge with data, elevating anomaly detection from simple data deviation analysis to a process with semantic interpretability. This paves the way for higher-level situational reasoning and causal analysis by combining knowledge graphs with large models. Chen et al. (Contribution 9) employ Bi-YOLO and OC-SORT algorithms for target ship recognition and tracking. This exemplifies the successful application of edge computing and lightweight AI models in real-time navigation assistance systems, which are crucial for achieving continuous “situation awareness” in autonomous navigation environments. Regarding dynamic and complex risk quantification, Chen et al. (Contribution 4) propose a dynamic calculation method for collision risk in complex navigable waters, where it essentially functions as a real-time decision-support system.

Building on accurate perception, the value of intelligent decision-making lies in prediction and optimization. This Special Issue demonstrates significant progress in ship motion prediction and autonomous control. In the realm of data-driven motion prediction and adaptive control, Guo et al. (Contribution 5) utilize LSTM networks to predict the maneuvering motion of Unmanned Surface Vehicles (USVs), while Wang et al. (Contribution 6) improve the Deep Deterministic Policy Gradient (DDPG) algorithm to optimize the PID controller for USV trajectory tracking. These two studies exemplify the integration of “AI + Control”: the former forecasts future states through temporal models, while the latter adaptively adjusts control policies via reinforcement learning to cope with complex sea conditions [3]. This marks a transition in ship control from traditional model-driven methods toward more resilient and self-learning, data-driven approaches.

In multi-agent collaboration and formation control, Tao et al. (Contribution 10) investigate cooperative formation control for multiple ships under time-delay conditions, and Han et al. (Contribution 3) adopt a multi-agent method to simulate ship movements in and out of the port in the Qiongzhou Strait waters. These studies extend intelligence from a single vessel to the coordinated optimization of ship groups. This represents a core characteristic for future intelligent shipping systems—leveraging algorithms like multi-agent reinforcement learning to achieve global optimal utilization of systemic resources such as ports and waterways. Optimizing the sequence of port entries/exits and berthing schedules can effectively reduce port turnaround times, providing a theoretical foundation for alleviating congestion and enhancing efficiency [4].

Intelligent shipping entails not only technological upgrades for individual ships but also the holistic reshaping of the entire maritime transportation system. Several articles in this collection, from a macroscopic systemic perspective, reveal the value of new technologies in enhancing system reliability and assessing external impacts. Han et al. (Contribution 8) apply percolation theory to analyze the reliability of the complex inland waterway transportation network in the Yangtze River. This is a typical big data system modeling approach capable of quantifying the impact of critical node or link failures on the entire network. Liu and Yu (Contributions 7 and 12) empirically analyze the impacts of nearshore and deep-sea offshore wind farm construction on maritime traffic complexity and safety, respectively. Such research fully leverages historical AIS big data, employing data mining and spatial analysis to quantify the interactions between human marine engineering activities and shipping operations. This provides critical decision support for future marine spatial planning and the synergistic development of “shipping-energy” systems.

This Special Issue also addresses emerging and sustainable topics such as polar navigation and alternative fuels. Liu et al. (Contribution 13) and Wang et al. (Contribution 14) investigate the influence of Antarctic sea ice and ship maneuvering in floating ice fields, respectively. These studies combine AI with physical models (e.g., coupled NDEM–MMG modeling) through numerical simulations to explore ship maneuverability in extreme, unstructured environments, accumulating valuable knowledge for the future development of polar routes. Zhou et al. (Contribution 2) analyze pathways for reducing the carbon emissions of ships in the Yangtze River using alternative fuels. Although not directly focused on AI, the findings from this work could serve as constraints embedded into AI-based fleet energy efficiency management systems, guiding ships to make economically and environmentally sound decisions in areas like route planning and speed optimization.