Search

Ship trajectory prediction is crucial for maritime trade and navigational safety. In this paper, we present a deep learning (DL) based trajectory prediction framework that can exploit the navigation pattern of a reference trajectory, a historical trajectory that resembles the target one, to enhance the prediction accuracy. A Differential Long Short-Term Memory (DLSTM) model is first proposed for trajectory prediction, which takes solely the past motion characteristics of the target trajectory as the input. Building on DLSTM, an enhanced DLSTM with reference trajectory correction (Ref-DLSTM) is proposed to integrate the features of both the target and the reference trajectory for better prediction accuracy. The DLSTM can be applied when a reference trajectory is not available, while the Ref-DLSTM is applied when a reference trajectory is present. To reduce the complexity of reference trajectory identification, a grid-based search algorithm is proposed to restrain the search in a local area. The efficacy of the proposed framework is evaluated using AIS datasets from the US Coast Guard and the Danish Maritime Authority. Numerical results demonstrate notable improvement over the state-of-the-art trajectory prediction methodologies, showcasing reductions in geographical prediction errors by 19.1% and 33.0% for DLSTM and 34.0% and 35.8% for Ref-DLSTM, respectively. • Deep-learning vessel trajectory prediction using DLSTM and Ref-DLSTM models. • Ref-DLSTM enhances predictions by utilizing reference and target vessel patterns. • Grid-based algorithm reduces complexity in matching target and reference trajectories. • Numerical results show superior prediction accuracy over state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Tropical cyclones frequently disrupt maritime transportation systems, impacting the normal operation of vessels and causing transportation delays. Analyzing the occurrence and degree of vessel transportation delays under disaster conditions is one of the critical aspects of maritime transportation management. This article, based on Automatic Identification System (AIS) data, examined the real behavioral trajectories of vessels and delay characteristics during tropical cyclones. Taking the example of Tropical Cyclone Veronica, which occurred in the waters off northwestern Australia in 2019, we conducted a comprehensive analysis. This involved numerical simulations of the entire disaster process and the cleaning and matching of trajectory data for all affected vessels in the area. We delved deeply into the relationship between vessel behavior characteristics and tropical cyclones, identified factors contributing to delays, and, based on our findings, constructed a Bayesian network inference model for vessel transportation delays under disaster conditions. This model uses information such as tropical cyclone intensity, vessel basic attributes, behavior choices, and port disaster avoidance measures as its primary nodes. The research results indicate that vessel rerouting, vessel loss of control, and waiting for port entry are the three major sources of vessel delays during disasters. Key influencing factors contributing to these consequences include port closure measures, the duration of vessels being adversely affected by strong winds, vessel tendencies toward loss of control, and risk preferences. To control the extent of vessel delays more effectively, it is crucial to prioritize these sensitive factors and strengthen monitoring and management efforts. This model is driven by real data, providing an objective reflection of real-world scenarios, and its effectiveness has been validated through sensitivity analysis. The research perspective and algorithmic framework presented in this paper offer a novel paradigm and fresh perspective for the study of maritime transportation disasters. The research findings have significant implications for bolstering the resilience of maritime transportation networks, enhancing our comprehension and anticipation of delays, and ensuring the continuity, security, and efficiency of the global supply chain for goods. • The maritime transportation delays resulting from tropical cyclones have been observed and given significant attention. • The delayed behaviors and characteristics of vessels under tropical cyclones are revealed by AIS. • A Bayesian network inference model for vessel transportation delays under tropical cyclones is constructed. • The key influencing factors of ship delay under tropical cyclones are identified. [ABSTRACT FROM AUTHOR]

Automatic evaluation of collision avoidance behavior with regards to safety and compliance with the International Regulations for Preventing Collisions at Sea (COLREGS) is a useful tool for developing safe and reliable collision avoidance algorithms, and a necessary component of any simulation-based verification scheme. As collision avoidance methods become more advanced, and capable of performing well in waters with restricted maneuvering space, accurate assessments of vessel behavior must also consider such restrictions. This paper presents methods for including considerations of the effect of grounding hazards on vessel behavior into an existing evaluation framework, originally developed for encounters in open waters. Encounters between vessels in normal operation, extracted from automatic identification system (AIS) data, are used to validate the evaluation methods. The validation is performed through employing both the original method and the extended method on each encounter. Selected scores and penalties from each method are then compared in order to display the effect of compensations given due to grounding hazards. The results show that grounding hazards preventing COLREGS compliant behavior does lead to compensation being given, also that the size of the compensation corresponds to the restrictiveness of the hazard. • This work incorporates considerations regarding grounding hazards in the evaluation of COLREGS compliance. • Trajectories of vessels are checked against ENCs to determine whether they were restricted by grounding hazards. • Different compensations are calculated for vessels with give-way and stand-on responsibilities. • The method can be employed on trajectories from simulations or collected from AIS. [ABSTRACT FROM AUTHOR]

Ship collision is a significant threat to waterborne transportation safety in busy waters. In order to mitigate the probability of ship collisions, improve navigation safety, and reduce economic losses, this paper proposes a spatial-temporal coupling approach to ship collision risk evaluation for alert in busy waters. First, the probabilistic ship domain considering ship characteristics (e.g., ship length and ship type) is determined based on historical Automatic Identification System (AIS) data. Secondly, the calculation models of the spatial-temporal collision risk of two-ship and multi-ship scenarios are proposed by the overlap of probabilistic ship domains, ship approaching trends, and catastrophe theory. Then, the ship probability domain overlap and the time required for ship collision avoidance are considered to estimate the threshold value of spatial-temporal collision risk between two ships. Meanwhile, a multi-level alert method for ship collision is proposed according to the degree of intrusion into the ship domain, the spatial-temporal collision risk, and the threshold value of the spatial-temporal risk collision. Finally, an example analysis was carried out based on the AIS data of the Luotou Channel in Zhoushan, China. The results indicate that the proposed method can effectively characterize the collision risk when two-ship or multi-ship encounters in busy waters, and can accurately classify the collision alert levels. Compared with traditional existing indicators, it is concluded that (1) the proposed method, which integrates both spatial and temporal collision risk, effectively categorizes collision alert levels, (2) it overcomes the limitations of traditional collision risk assessment methods, particularly in reflecting nonlinear changes in collision risk, and (3) it can assess collision risk in busy waters. • Develop a spatial-temporal coupling approach to chip collision risk evaluation for alert in busy waters. • Propose an AIS-based probabilistic ship domain method divided into the forbidden area and risk area. • Verify the model using real-world AIS data from the Luotou Channel in Zhoushan, China. • Contribute to relieving regulators and crew of the challenges of relatively high traffic density or complexity. [ABSTRACT FROM AUTHOR]

Traditional risk analysis models often struggle to provide effective insights into real-time and objective risk assessment of ship navigation. Therefore, this paper introduces a ship navigation risk online prediction model (NROPM) based on the fusion of multi-source data. Firstly, this paper extracts risk feature information from multiple sensors, such as AIS data, wind, and water currents. These data are integrated through feature fusion to create the research dataset. Subsequently, the entropy weight-cloud model is constructed to calculate the risk and generate a set of test samples. An emerging informer network is introduced to train the model and predict the risk of ship navigation. Finally, this paper compares the proposed model with traditional neural network models through comparative experimental. The research was conducted using a real dataset from the Pearl River area over a period of three months. Model performance was assessed in terms of mean squared error and training time consumption. The experimental results indicate that the proposed NROPM outperforms traditional models like the BP (backpropagation) neural network and transformer in both accuracy and computational efficiency for real-time ship navigation risk prediction. • This paper proposes a ship navigation risk online prediction model using multi-source data. • An improved informer network is used to achieve the real-time risk prediction. • A quantification method for ship navigation risk is introduced, and the accuracy of the evaluation results is verified using entropy-weighted cloud models. • Experimental results show this model's superior accuracy over traditional methods, enabling objective, dynamic ship navigation risk assessment in busy waters. [ABSTRACT FROM AUTHOR]

Ports are essential facilities for the functioning of our globalised economy, as well as for the territories they supply. Monitoring port activity is therefore a relevant way of assessing the economic intensity they generate, and enabling regional analyses of their relative weight, their areas of activity and their development trajectories. However, comparative analysis of port activity is hampered by the variable quality and uneven distribution of the data produced by port authorities, which is far from being homogeneous, which makes comparative approaches difficult. In this paper, using vessel positioning data obtained through the Automatic Identification System (AIS), we propose a method for the extraction, quantification and qualification of port calls. We apply this method to a dataset in the Caribbean, covering the whole of 2019. Our method was able to describe the activity of 101 ports (in the sense of port areas) in the Lesser Antilles, by identifying 171,626 calls and their duration, by 5,907 ships of various type, size, age or flag. A comparison with the statistics published by a number of reference ports authorities and by the UNCTAD (United Nations Conference on Trade and Development) shows that the median difference of computed port calls is about 11%, thus validating the proposed approach. By qualifying the type of vessel that passes through port waters, according to its characteristics, this approach based on the number of calls makes it possible to observe the capabilities and the infrastructural quality deployed in the ports in terms of vessels, and, stemming from it, to develop the hypothesis of port vulnerability, that is a perspective for our future work. • A method to extract port calls from AIS data has been developed and implemented. • A careful survey of all quays in the Lesser Antilles allows a great degree of details. • Harbour activity can be characterised by the diversity in the types of vessel calling. • Port call data such as duration or ship age and tonnage enhance traffic understanding. • The volume of computed port calls is validated by comparison against official data. [ABSTRACT FROM AUTHOR]

This research paper proposes two different methods for removing biases from geo-spatial trajectories obtained from AIS, enabling the creation of machine learning models that generalize well to areas with no training data. We focus here on the task of behavior detection such as moored, underway or drifting from AIS. The first method utilizes data augmentation techniques specifically designed for augmenting geo-spatial trajectories that include angled information such as direction, while the second method is based on engineering features that eliminate these geographical biases. These two methods are compared using different types of machine learning models, including random forests, hidden Markov models, LSTMs, and transformer networks. • Vessel behavior detected from AIS with machine learning models. • Data augmentation or engineering of unbiased features allow us to generate global models. • Comparison of different models for behavior detection such as HMM, LSTM and transformer networks. [ABSTRACT FROM AUTHOR]

This paper presents a correlation between the gaseous pollutant concentration measured in port and the estimated in-port ship emissions based on AIS data for the port of Naples. With the aim of better monitoring the quality ship's plumes funnels, the concentrations of Nitrogen dioxide (NO 2) and Sulphur dioxide (SO 2) have been continuously measured, during an experimental campaign (May–June 2021) at a height of 20 m from ground level. An innovative bottom-up ship emission modeling system with an emission inventory was created; five principal steps were realized: acquisition and cleaning, construction of ship's database, analysis of ship routes and phases in port, calculation of fuel consumption, and emission. The updated methodology bypasses the typical fixed value of specific fuel consumption, which is obtained based on the engine type and fuels and uses a correlation with the real engine load. Several updates were adopted for the emission estimation, abandoning the constant emission factor with the use of a more detailed methodology. The correlation was obtained by taking into account the prevailing winds and the overlap between the experimental campaign and the emission inventory on particular wind sectors. As a results correlations between the pollutant flow rates estimated from the AIS data and the concentrations measured during the experimental campaign, have been find (0.97 for SO 2 and 0.80 for NO X). • Experimental campaign in port. • Emission inventory based on AIS data. • Overlap between full-scale measurement test in port and emissions estimation based on AIS data. [ABSTRACT FROM AUTHOR]

A weighted track fusion algorithm is proposed to address the complexity of conventional sea monitoring data and the lack of effective and convenient monitoring methods. The new algorithm uses local information entropy and fuses data from the automatic identification system (AIS) and X-band radar. The new model employs ship monitoring data to detect five types of abnormal ship behavior: speed, course, position, ship spacing, and time abnormalities. The method weights various factors using a weight matrix built up from a distance difference matrix between different track points and calculations of the distance information entropy and residual degree using information entropy theory. We tested the proposed weighted track fusion algorithm the fusion accuracy using actual data for 2813 ships, AIS, and radar track points based on local information entropy. The resulting correlation accuracy was 95.54% for 2716 ship track points and a correlation accuracy of 90.72% for 97 ship tracks. The results indicate that the proposed algorithm is suitable for fusing multiple tracks and is more robust than traditional fusion methods. • A weighted track fusion algorithm is developed to address the complexity of conventional sea monitoring data and the lack of effective and convenient monitoring methods. • The new algorithm is based on local information entropy and the fusion of AIS and X-band radar data. • This paper proposes an innovative algorithm that compensates for incorrect speeds, headings, positions, and ship spacing; the accepted but time-abnormal mathematical model is also corrected. [ABSTRACT FROM AUTHOR]

The slow convergence, excessive turning points and non-smooth path generation are significant challenges in ship path planning by existing RRT-related algorithms. To address these issues, this paper proposes a novel approach called PI-DP-RRT, which combines prior automatic identification system (AIS) information and Douglas-Peucker (DP) compression for ship path planning. Firstly, we cluster the available AIS data to construct guide regions, the guidance region can provide information to guide the target in the RRT algorithm. Next, we improve the algorithm's sampling strategy based on the guide region by using paranoid sampling, which increases convergence rate. Finally, we optimize the path by applying the improved DP algorithm and a new path optimization method to enhance its smoothness and practicability. Comparative simulation experiments are conducted in real scenarios and challenging environments. The results indicate that the proposed PI-DP-RRT algorithm outperforms other RRT-related algorithms in terms of efficiency, achieving a good balance between algorithm efficiency and accuracy. The planned path reduces the turning range and improves the smoothness of the path, which promotes the safety and efficiency of ship navigation. • This paper proposes an improved RRT algorithm based on AIS information for ship path planning. • The proposed method takes AIS data as a priori information. • Integrate DP and turning optimization algorithm to make the planned path more practical. • The proposed PI-DP-RRT algorithm effectively balances algorithm efficiency and accuracy. [ABSTRACT FROM AUTHOR]

Vessel trajectory prediction plays a vital role in maintaining a safe and effective status in maritime transportation. The development of deep learning provides appropriate mechanisms to predict vessel trajectory using Automatic Identification System (AIS) data. However, besides limited historical position information, much information can be used to enhance deep models, including speed, course, departure time, sailing distance and vessel type. Realistically speaking, these are related to the ships' endurance, oil storage, tide and work–rest habits. In order to address these problems, we propose a novel vessel trajectory prediction model METO-S2S based on a sequence-to-sequence structure, consisting of the Multiple-semantic Encoder and the Type-Oriented Decoder. After comprehensive data preprocessing and feature engineering, the Multiple-semantic Encoder embeds the sequential inputs of the voyage information into higher-dimensional latent vectors. In addition, the combined vector of vessel type and departure time of a particular ship is fed into the Type-Oriented Decoder treated as guidance information. To verify the efficiency and efficacy of METO-S2S, we employ an AIS dataset in US coastal waters which is more suitable for deep learning models, and implement comparative experiments with several baseline models. The experimental results show that METO-S2S is superior to them both quantitatively and qualitatively. 1 1 The code and data of our experiments is available on GitHub https://github.com/AIR-SkyForecast/METO-S2S. • The paper releases a toolkit for the AIS data preprocessing and a processed AIS dataset of US coastal waters which is more suitable for deep learning methods. • The paper develops a novel long-term vessel trajectory prediction method METO-S2S based on a Seq2Seq structure which makes good use of track information and diversified sailing features. • The paper proposes a Multiple-semantic Encoder that combines coordinate, speed, course, and sailing distance characteristics. Furthermore, the paper also proposes a Type-Oriented Decoder taking the combined vectors of vessel type and departure time as guidance information. • METO-S2S is evaluated on our released trajectory dataset. Compared with other baselines, METO-S2S can generate more accurate and robust prediction results. [ABSTRACT FROM AUTHOR]

Multi-vessel collision risk assessment for maritime traffic surveillance is a key technique to ensure the safety and security of maritime traffic and transportation. This paper proposes a framework of real-time multi-vessel collision assessment that combines a spatial clustering process (DBSCAN) for detecting clusters of encounter vessels and a multi-vessel collision risk index model for encounter vessels within each cluster from the large amounts of monitored vessels in a surveyed sea area. First, the vessels monitored are clustered using DBSCAN to obtain the clusters of encounter vessels, filtering out the relatively safe vessels. Then, the dynamic motion relation between encounter vessels within each cluster is modeled to obtain DCPA and TCPA. The semantic and mathematical relationship of vessel collision risk index for each cluster of encounter vessels with DCPA and TCAP is constructed using a negative exponential function. To illustrate the effectiveness of the framework proposed, an experimental case study has been carried out within the west coastal waters of Sweden. The results show that our framework is effective and efficient at detecting and ranking collision risk indexes between encounter vessels within each cluster, which allows an automatic risk prioritization of encounter vessels for further investigation by operators. Hence, this framework can improve the safety and security of vessel traffic transportation and reduce the loss of lives and property. [ABSTRACT FROM AUTHOR]

Atmosphere pollution brought by rapid development of shipping industry becomes increasingly serious, and has brought great challenges to maritime supervision and management. With the popularization of AIS (Automatic Identification System) technology, massive navigation logs recording rich ship activities can be collected. These logs make it possible to apply a data-driven method to quantify exhaust emissions from ships. This paper is therefore motivated to research methods of approximately estimating ship exhaust emissions and exploring its spatial distribution and variation rules. Firstly, ship static and dynamic information extracted from AIS logs are utilized to distinguish different ship navigational states. A quantitative model is then created to estimate ship emissions from main engine, auxiliary engine and boiler with varied work conditions in different states. Spatio-temporal analysis methods are further applied to comprehensively discovery knowledge hidden in ship emission inventories. Typical analyses from space, time and attributes have uncovered much useful knowledge for ship emission supervision, such as high-polluted area, peak-emission time, least environmentally friendly ship type, etc. Finally, Ningbo-Zhoushan port plays a role as case study in validating the proposed method. The experiment results illustrate this research can be helpful to make efficient emission control measures and improve the capability of environmental management. [ABSTRACT FROM AUTHOR]

Ship collision risk analysis is of great significance for maritime traffic management and surveillance in real operational conditions. However, the traditional concept of the closest point of approach (CPA) is limited, particularly in coastal areas. Thus, this paper introduces a quantitative analysis method for measuring ship collision risk. The proposed method uses both the static and dynamic information of Automatic Identification System (AIS) data. First, the closest points of ship-ship collision (CPC) are calculated based on the ship specifications (e.g., ship length and ship breadth) and the geographical positioning of ships. Dynamic ship collision boundaries are estimated via CPC for the involved ships. Then, a kinematics feature-based vessel conflict ranking operator (KF-VCRO) is introduced to evaluate ship collision risk by integrating the relative position vector and the relative velocity, accounting for static and dynamic information of AIS. Finally, our method is validated by simulating head-on, overtaking, and crossing situations, showing that it can accurately assess ship collision risk for typical ship collision scenarios. Especially, we further validated our proposed method through real-world experiments in Zhoushan, China. The results indicate that the proposed method provided accurate identification of high collision risk areas in Zhoushan in real maritime practices. It is concluded that (1) the method assists in quantifying ship collision risk and identifying high collision risk areas, and (2) estimating traffic risk provides further insight into maritime traffic surveillance. • A quantitative analysis method is proposed to quantify collision risk. • The method makes use of static information and dynamic information from AIS. • Time-varying motion trajectory model is proposed to present ship dynamics. • The proposed method is useful for high-risk hotspots identification. [ABSTRACT FROM AUTHOR]

Trajectory prediction is a critical technology for ensuring ship navigation safety, improving the efficiency of marine traffic control, and efficiently searching for maritime targets. This research proposes an Improved Social-STGCNN (IS-STGCNN) to increase the accuracy of ship trajectory prediction. This technique successfully combines the benefits of the spatio-temporal graph convolutional neural network (STGCNN) for extracting ship trajectory features and the benefits of the time-extrapolated convolutional neural network (TXP-CNN) for generating future ship trajectory using the obtained features. This method proposes a social-sampling strategy based on prior knowledge, i.e., the bumper model of ship safety assessment, to carry out negative sampling and enhances the effectiveness of the IS-STGCNN method in learning the concept of negative examples (such as collisions). The interaction force between ships is estimated in a novel way in the spatio-temporal graph convolutional neural network to improve high-precision feature extraction from ship trajectory. Model predictive control (MPC) is used to correct the output of trajectory prediction in order to fulfill kinematic constraints and assure prediction kinematic feasibility. The experiments are carried out using AIS data, and the findings reveal that IS-STGCNN outperforms the state-of-the-arts. • The Improved Social-STGCNN (IS-STGCNN) is proposed to increase the accuracy of ship trajectory prediction. • The Social-sampling module is utilised to update the vector representation of nodes for better feature extraction. • The suggested method uses Model Predictive Control (MPC) module to ensure the kinematic feasibility of the prediction. [ABSTRACT FROM AUTHOR]

Automatic identification systems (AIS) are on-board compulsory devices on ships that can provide a large amount of dynamic vessel information, such as position, course and speed information, which is meaningful for marine management. This paper explores the potential to utilize AIS data to recognize ship manoeuvring hot zones and analyse their characteristics to improve marine management efficiency. First, considering manoeuvring behaviours such as turning, acceleration and deceleration, a novel approach based on MiniBatchKMeans that recognizes ship turning hot zones, acceleration hot zones and deceleration hot zones is proposed according to course and speed variation. Second, joint probability density functions for speed over ground (SOG) and course over ground (COG) are built taking into account the combined influences of motion characteristics, which are beneficial for determining the ship traffic rules in manoeuvring hot zones and utilized in marine management. Finally, numerical and comparison experiments are conducted to test the feasibility and effectiveness of the proposed method. • The designed recognition algorithm can identify manoeuvring hot zone based on speed and course variations. • The established joint probability density distribution model is able to uncover ship traffic rules in manoeuvring hot zones. • The proposed methodology is effective in marine management. [ABSTRACT FROM AUTHOR]

The universality of the automatic identification system (AIS) as well as the electronic chart display and information system (ECDIS) installations will provide revolutionary solution schemes for the e-navigation era of informatization, intellectualization, and integration. Based on this concept, this study adopted ECDIS as an information platform for navigational decision support and used the real-time navigation information received by the AIS to construct predicted areas of danger (PAD) for target ships. The advantages of direct viewing via the PAD in the collision avoidance give PAD a new application. Subsequently, spatial data from an electronic navigation chart (ENC) was employed as a basis for generating geographical obstacles. Through the integration of a geographic information system (GIS) module, specially designed evolutionary computation, and collision prevention regulations (COLREGS) knowledge as well as by comprehensively considering overall navigation situations, this system conducted obstacle avoidance processing and selection of a route. The system could generate a route that could simultaneously achieve multi-ship encounter collision avoidance and geographical obstacle avoidance. This is expected to provide mariners with greater convenience when working at sea, reducing their workload. This system can also be used as a reference for collision avoidance decision making. [ABSTRACT FROM AUTHOR]

This study characterizes navigation conflicts in a region with a large traffic volume along the United States (US) Atlantic Coast, using an empirical approach that analyzes Automated Identification System (AIS) data for a full year (2010). The region includes areas proposed for wind energy development, and this characterization of vessel conflicts could be useful as a baseline to evaluating the effect of offshore wind areas on navigation conflicts and vessel collisions. There have been no collisions reported within the study area. This study analyzes the one year of AIS data to obtain vessel tracks and makes simultaneous pairwise comparisons to determine the number and rate of vessel encounters and near-misses (navigation conflicts). For vessel pairs in an encounter (3 nm distance; crossing, head-on, overtaking), a 'blind vessel' assumption was used to calculate the number of potential near misses and potential collisions. Graphical analysis of the frequency of vessels involved in potential near misses produced vessel response curves that provided the time of peak frequencies prior to the time of closest approach for each encounter type. These peak times were up to 600 s prior to the time of closest approach. The analyses of potential near misses and collision used the data during the period between the time of peak frequency and the time of closest approach to filter the data used in the calculation of the probability of potential collisions. The period included collsion avoidance maneuvering. Statistical analyses are made to estimate the point values and uncertainty of near miss rates and the probability of potential collisions for each type of encounter from which bounding estimates of the collision rates are made. These bounding collision rates are found to be comparable with estimates from previous studies. Causation probabilities were also estimated and compared with previous studies. The response curves may also have utility in the simulation of navigation conflicts. The use of surrogate data to evaluate the effect of wind energy areas is considered, though their usefulness depends on comparability of traffic density and the level and type of traffic control systems implemented. • Evaluates the occurrence of encounters and near-misses along the US Atlantic Coast that include proposed wind energy areas. • Examines the overall vessel responses to potential near-misses. • Estimates the bounding rate collisions and causation probabilities in the analysis domain. • Discusses the possible impacts of the presence of wind energy areas on collision rates. [ABSTRACT FROM AUTHOR]

Seamen usually select popular routes according to navigational and hydrographical characteristics when passing through a strait with congested traffic. In order to minimise the possibility of collisions, Principal Fairways (PFs) are commonly used by sailors. It is essential to delineate PFs quantitatively and objectively for designing or refining routing measures. In this paper, a space use method found in habitat evaluation of wildlife is applied to extract PFs of ship passages through a strait. Compared with existing methods, the proposed method helps to identify cumulative activity patterns for ship groups derived from mass ship trajectories, and provides a clearer interpretation of shifting space-use patterns within strait corridors. Moreover, it gives a better insight for directional and seasonal factor for PFs in straits. Finally, this novel method is used to extract PFs in western Taiwan Strait and its adjacent sea. The results indicate that the proposed method is helpful to identify gaps between current ship routing system plan and cumulative activity patterns recognised by real ship trajectories. [ABSTRACT FROM AUTHOR]

The AIS is mostly used to exchange different vessel data between vessels and AIS base stations. For mariners AIS data have become a useful complement to radar data since manual collision avoidance is supported by providing additional information via AIS. Furthermore AIS allows for implementing course prediction algorithms. However, AIS message losses have been confirmed by several studies affecting and restricting the AIS usage especially within vessel prediction algorithms. Several evaluations and simulations concerning the Self-Organized Time Division Multiple Access (SOTDMA) algorithm show that an AIS message loss caused by system overload is unlikely. However, in contrast to the SOTDMA evaluations there are no evaluations about the influence of the VHF medium and the AIS antenna setup on the AIS reception rate. Therefore, we intend to identify the VHF influence on the reception rate by using high quality AIS hardware and AIS antenna installations according to IMO guidelines to record and evaluate AIS data on a German Search and Rescue vessel. By using two different antennas we evaluate the meaning of the antenna height as well as of the general antenna setup for the AIS message reception rate and show the impact of measured reception rates for vessel prediction algorithms. [ABSTRACT FROM AUTHOR]

AIS (Automatic Identification System) data provides valuable input parameters in ship traffic simulation models for maritime risk analysis and the prevention of shipping accidents. This article reports on the detailed comparisons of AIS data analysis between a Dutch case and a Chinese case. This analysis focuses on restricted waterways to support inland waterway simulations, comparing the differences between a narrow waterway in the Netherlands (the Port of Rotterdam) and a wide one in China (wide waterway of Yangtze River close to the Su-Tong Bridge). It is shown that straightforward statistical distributions can be used to characterize lateral position, speed, heading and interval times for different types and sizes of ships. However, the distributions for different characteristics of ship behaviors differ significantly. [ABSTRACT FROM AUTHOR]

The worldwide increase in ship traffic poses a series of challenges for the designer, the crew and the authorities involved in ship traffic management. Depending on the ship type, both onboard comfort and safety represent two key aspects to be taken into account during the ship life. In the present paper, the focus is on the assessment of the safety and comfort in service during navigation in real weather conditions by focusing on an entire route instead of on a single ship. A method for assessing the safety and comfort levels of specific routes is presented and discussed in the paper. In particular, a statistical approach to set up limit thresholds able to quantify the level of navigation safety and comfort for different ship types in a specific area is presented based on historical weather data. The method is based on the Automatic Identification System (AIS) data for the traffic characterization, a strip theory for the seakeeping calculations and historical meteorological data for the environment definition for a specific area in the Mediterranean Sea. • Definition of the statistical range of safety and comfort indexes for ships in the Mediterranean sea area. • Use of the AIS and weather data in the Pelagos area to extract real safety navigation indexes. • Definition of a robust and reliable procedure to define safety and comfort indexes for ships in a sea region. [ABSTRACT FROM AUTHOR]

Structures exposed to cyclic wave loading often develop fatigue cracks, demanding maintenance efforts. Ships are designed to withstand the same wave environment independently of their operations. This paper investigates the operational patterns of five different groups of vessels, focusing on the structural fatigue effects. The investigation combines almost 100 million hours of worldwide Automatic Identification System (AIS) data, environmental hindcast data, and simplified structural models. Results show that vessel operators change behavior in rough waves (typically from Beaufort scale 6 for smaller vessels), though speed and heading changes depend on vessel size and type. These changes result in more fatigue damage accumulation than calculations using the common assumption of zero speed and uniform heading probabilities. Consistent with previous studies limited to North Atlantic traffic, worldwide, ship operators make choices that mitigate ships' encountered waves. • Extensive worldwide vessel transit and hindcast data quantify storm avoidance effects. • Combined wave and transit data describe operator actions in various wave conditions. • Varied rough-weather operation between vessel groups suggest effects of type and size. • Simple fatigue calculations show conservatism of standard assessment wave statistics. • Enriching ship tracks with wave data gives representative operational profiles. [ABSTRACT FROM AUTHOR]

Ship trajectory clustering is one of the main methods of ship trajectory mining based on AIS data. However, there exist two main problems in trajectory clustering: One is the inherent problem of the clustering algorithm, such as low computational efficiency, poor recognition of noise, sensitivity to the density distribution of data, etc.; the other problem is about the similarity measurement of the trajectories, that is, most of the measurement methods ignore the overall motion trend or local features of ship trajectories. In order to solve these problems, this paper proposes the FOLFST, a novel ship trajectory clustering method for Finding Overall and Local Features of Ship Trajectories, which consists of four parts: 1) segment the trajectory with the ADP (Adaptive-threshold Douglas-Peucker) algorithm; 2) adopt new sub-trajectory segments similarity measurement method which considers both the wholeness and locality of the trajectory; 3) cluster the sub-trajectory segments with the improved FOP-OPTICS (Finding of the Ordering Peaks Based on OPTICS); 4) identify sub-trajectories which belong to multiple clusters. The experimental results show that the FOLFST has obvious advantages over the other four algorithms in identifying the overall and local features of ship trajectories, noise recognition, computational efficiency, etc. • Our method solves the shortcomings of traditional clustering algorithms in parameter setting and noise recognition. • The FOLFST can find the overall movement trend of trajectories, the local abnormal segments and local similar features. • The FOLFST solves the problem of exclusivity of traditional algorithm. • The gridding method improves the computational efficiency of similarity measurement. [ABSTRACT FROM AUTHOR]

To explore the regularity of ship density and ensure the safety of maritime traffic, a ship density model based on ship scale conversion and grid is established by considering the impact of ship size on traffic safety. A non-parametric kernel density estimation (KDE) method is used to obtain the probability density function of the ship density by selecting the appropriate kernel function and determining the optimal bandwidth. To process and store Automatic Identification System (AIS) data efficiently, a grid-based storage method of AIS data is proposed in this paper, and the relationship between grid width and sampling interval is established. The experiments result proves that the grid-based AIS data storage method saves 64% of storage space compared with the original AIS data. The ship density model based on ship scale conversion can reflect the impact of ship size on the maritime traffic. The distribution regularity of ship density can be obtained through the probability density curve of ship density and statistical characteristics. The research results show that the proposed method can provide theoretical support for the safety supervision of the maritime authority. It can also be recommended for the evaluation of VTS centers in vessel traffic management. • A ship density calculation model based on ship scale conversion is established to reduce the differences caused by different ship sizes. • A grid-based storage method for AIS data is proposed to improve processing efficiency. • A non-parametric KDE method is used to obtain the probability density function of ship density by selecting the appropriate kernel function and the optimal bandwidth. • The distribution characteristics and law of ship density in different waters are analyzed based on probability density curve and statistical characteristics of ship density. [ABSTRACT FROM AUTHOR]

Ship trajectory information has made a significant contribution to the data-based research in analyzing maritime transportation and has facilitated the improvement of maritime safety. However, the AIS data, which consists of ship trajectory, inevitably contains noises or missing data that can interfere with the conclusion. In this paper, an improved kinematic interpolation is presented for AIS trajectory reconstruction, which integrates data preprocessing and interpolation that considers the ships' kinematic information. The improved kinematic reconstruction method includes four steps: (1) data preprocessing, (2) analysis of time interval distribution, (3) abnormal data detection and removal, (4) kinematic interpolation that takes the kinematic feature of ships (i.e., velocity and acceleration) into account, adding forward and backward track points to help correct the acceleration function of reconstruction points. The proposed method is tested on the AIS dataset of Zhoushan Port and was compared with traditional ship trajectory reconstruction methods. The comparison indicates that the proposed method can effectively reconstruct the ship trajectory with higher performance on a single ship trajectory and a large AIS data set of certain water areas. • Comprehensive and detailed data processing is integrated into trajectory reconstruction. • Kinematic interpolation is improved considering harmonic points to refine the reconstruction. • The proposed method applies to large-scale data and has high performance. [ABSTRACT FROM AUTHOR]

The deficiencies of massive data like storage difficulty, computation inefficiency, and information redundancy call for ship trajectory compression. While most studies on ship trajectory compression have one or more of the following drawbacks: the drawback of low compression efficiency; the problem resulted from error ship static information when the distance threshold is set based on ship length or width; poor compression quality for some trajectories, which is caused by experience-based optimal threshold. To solve these problems, we propose the ADP (Adaptive-threshold Douglas-Peucker) algorithm based on DP (Douglas-Peucker) algorithm. By determining the key points of each trajectory through the threshold change rate, ADP no longer relies on ship static information and makes it easier to determine the threshold, which is what traditional algorithms cannot achieve. Additionally, we use the advantage of matrix operation and the method of reducing points to improve the algorithm's computation efficiency. To verify the feasibility and superiority of the proposed algorithm, we compared our algorithm with DP algorithm, Partition-DP algorithm and Sliding Window algorithm from four aspects, namely, compression rate, Synchronous Euclidean distance, Length Loss Rate and running time. The experimental results prove that our algorithm has advantages over the other three algorithms, especially in threshold setting. • The proposed algorithm solves the problem resulted by the experience-based threshold. • It solves the problem of poor or even unsuccessful compression due to incorrect ship static information. • We use the matrix operation and the method of reducing points to improve the algorithm's computation efficiency. [ABSTRACT FROM AUTHOR]

The intense maritime traffic of modern days requires that a considerable attention be paid toward the comfort of ships not only as design criteria, but also as a means of assessing the quality of navigation in service conditions. Certain passenger routes and ship dimensions are more vulnerable than others to rough weather and the current climate changes require adaptation of design and navigation warning strategies. The main aim of the paper is to describe a method to derive a probabilistic representation of the comfort of passengers and crew on a selected route. AIS data is used to identify and characterise the chosen route. The seakeeping properties of the principal ships are evaluated employing a strip theory code. Detailed wave hindcast data is then employed to build state-of-the-art wave scatter diagrams along the route which allow the calculation of the ship responses for each sea state and the associated probability. The Motion Sickness Incidence is finally used as a comparative measure of the overall on-board comfort experienced on a variety of passenger vessels. • A simple statistical analysis of real AIS data is employed to identify and characterise a large passenger route. • A strip theory program is used to assess the seakeeping properties of representative ships of the selected route. • State-of-the-art wave scatter diagrams are generated from detailed marine weather hindcast data. • The probabilistic assessment of passenger comfort is carried out on the route combining seakeeping characteristics and probabilistic weather data. • Motion Sickness Incidence is employed as a relative measure of the overall n-board comfort level. [ABSTRACT FROM AUTHOR]

The ship handling decision-making process constitutes the most critical step in intelligent ship collision avoidance. This study utilized the ship encounter azimuth map to extract the data of 12 types of ship encounter modes from the automatic identification system (AIS) big data. The sliding window algorithm was used to identify the ship encounter behaviors, which were used as the training data. The sequence conditional generative adversarial network (Seq-CGAN) was proposed using the sequence-to-sequence (Seq2Seq) model to learn how to generate appropriate anthropomorphic collision avoidance decisions and bypass the process of ship collision risk assessment based on the quantification of a series of functions. The long short-term memory (LSTM) cell was simultaneously combined with Seq-CGAN to improve the memory capacity and the current adaptability of the environment of the overall system. The Seq-CGAN was trained by 2018 Zhoushan Port AIS data to achieve structured learning regarding human experiences. The results indicate that the Seq-CGAN effectively formulates the sequence of ship handling behaviors. This study contributes significantly to the increased efficiency and safety of sea operations. The proposed method could be potentially applied to determine the predictive framework for various intelligent systems, including intelligent collision avoidance, route planning, and operational efficiency estimation. • Proposed the application of multiple consecutive SHBB sequences as a new representation method for ship handling decision-making. • Construct the Seq-CGAN model based on Seq2Seq learning the underlying sequence of handling behaviors implicit in AIS data. • Combining LSTM cell with Seq-CGAN to extend Seq-CGANs into deep neural networks, enhancing the stability and memory of the entire model. • Proposed the 7-tuple coding and Capsule structure coding transformation to fully express a ship handling behavior unit and a recommended splicing threshold is given. [ABSTRACT FROM AUTHOR]

Advances in artificial intelligence are driving the development of intelligent transportation systems, with the purpose of enhancing the safety and efficiency of such systems. One of the most important aspects of maritime safety is effective collision avoidance. In this study, a novel dual linear autoencoder approach is suggested to predict the future trajectory of a selected vessel. Such predictions can serve as a decision support tool to evaluate the future risk of ship collisions. Inspired by generative models, the method suggests to predict the future trajectory of a vessel based on historical AIS data. Using unsupervised learning to facilitate trajectory clustering and classification, the method utilizes a cluster of historical AIS trajectories to predict the trajectory of a selected vessel. Similar methods predict future states iteratively, where states are dependent upon the prior predictions. The method in this study, however, suggests predicting an entire trajectory, where all states are predicted jointly. Further, the method estimates a latent distribution of the possible future trajectories of the selected vessel. By sampling from this distribution, multiple trajectories are predicted. The uncertainties of the predicted vessel positions are also quantified in this study. • This study applies machine learning techniques to historical AIS data. • A dual linear autoencoder approach is introduced, inspired by deep generative models. • The approach predicts the future trajectory of a vessel, to aid in collision avoidance. • A procedure to evaluate the uncertainty of the trajectory prediction is also presented. • The approach is evaluated using test cases from the region of Tromsø, Norway. [ABSTRACT FROM AUTHOR]

Ship collision is one of the major contributors of maritime accidents. Quantitative risk analysis of such accidents is an effective tool for maritime safety administrations to understand the current risk level and propose risk mitigation measures. In this paper, an improved Time Discretized Non-linear Velocity obstacle (TD-NLVO) algorithm is proposed to detect multiple ship encounter situations using historical AIS data. Boolean operation on the individual NLVO is integrated with TD-NLVO using the union of the velocity-obstacle sets to determine a dangerous encounter situation according to the pre-set criteria. Two case studies are implemented to illustrate the capability of the proposed algorithm. A comparison is conducted between the previous and the improved methods. The results indicate that the improved method can effectively identify a multiple ship encounter which satisfies the pre-set criteria. The improved method has the potential to provide more detailed information for stakeholders e.g. maritime safety administration, etc. to propose risk mitigation measures as well as to improve the accuracy of geometric probability analysis for ship collision risk. • New approach for multiple ship encounter analysis using Non-Linear Velocity obstacle and AIS data. • Individual VOs are combined the construct the velocity obstacles for multi-ship encounter. • TD-NLVO is introduced to build individual velocity obstacle set induced by each target ship. [ABSTRACT FROM AUTHOR]