Spatio-Temporal Deep Learning Approaches for Addressing Track Association Problem using Automatic Identification System (AIS) Data

Maritime vessel movement prediction: A temporal convolutional network model with optimal look-back window size determination

The Automatic Identification System (AIS) has recorded near-real-time vessel monitoring data over the years, paving the way for data-driven maritime surveillance methods; concurrently, the data suffer from unrefined, reliability issues and irregular intervals. In this paper, we address the problem of vessel destination estimation by exploiting the global-scope AIS data. We propose a differentiated data-driven approach recasting a long sequence of port-to-port international vessel trajectories as a nested sequence structure. Based on spatial grids, this approach mitigates the spatio-temporal bias of AIS data while preserving the detailed resolution of the original. Further, we propose a novel deep learning architecture (WAY) that is designed to effectively process the reformulated trajectory and perform the long-term estimation of the vessel destination ahead of arrival with a horizon of days to weeks. WAY comprises a trajectory representation layer and channel-aggregative sequential processing (CASP) blocks. The representation layer produces the multi-channel vector sequence output based on each kinematic and non-kinematic feature collected from AIS data. Then CASP blocks include multi-headed channel- and self-attention architectures, where each processes aggregation and sequential information delivery respectively. Then, a task-specialized learning technique, Gradient Dropout (GD), is also suggested for adopting many-to-many training along the trajectory progression on single labels. The technique prevents a surge of biased feedback by blocking the gradient flow stochastically using the condition depending on the length of training samples. Experimental results on 5-year accumulated AIS data demonstrated the superiority of WAY with recasting AIS trajectory compared to conventional spatial grid-based approaches, regardless of the trajectory progression steps. Moreover, the data proved that adopting GD in a spatial grid-based approach leads to the performance gain. In addition, the possibilities of improvement and real-world application with WAY's expandability in multitask learning for the estimation of ETA was explored.

Maritime crimes such as Illegal, Unreported, and Unregulated (IUU) fishing, piracy, and smuggling pose significant threats to marine ecosystems, trade, and coastal security, especially in developing regions. Automatic Identification System (AIS) data offers a scalable solution for vessel monitoring, but the use of fully supervised machine learning models is constrained by the substantial manual effort and expert input required to label training data. To address this challenge, authors propose a semi-supervised machine learning pipeline that classifies vessel activities from AIS data without relying on pre-labeled datasets. Our approach leverages scaled geospatial and temporal features, including latitude, longitude, speed, and time difference, to train multiple Hidden Markov Models (HMMs) on trajectory segments. These segments are then grouped using similarity-based K-means clustering and subsequently classified with supervised models, including Random Forest and Long Short-Term Memory (LSTM) networks. The pipeline effectively identifies and labels maritime activities such as sailing, fishing, idling, and other activities. Experiments were conducted on a dataset comprising 156,379 AIS points, partitioned into training and test sets. The LSTM-based supervised model achieved an F1 score of 0.86 on the local test set, while the end-to-end pipeline achieved an F1 score of 0.5 on a global evaluation set. These results demonstrate the feasibility of automating maritime activity classification through artificial intelligence and hybrid learning, offering a scalable solution for real-world maritime surveillance.

Maritime traffic pattern extraction is a fundamental and crucial factor for maritime surveillance and anomaly detection. Emerging technologies like Automatic Identification System (AIS) provides multi-dimensional data which is used to construct a maritime traffic model. In this paper, we propose a framework of maritime traffic pattern extraction from vessel AIS information, which learns a traffic pattern using an unsupervised technique, and can be applied on historical Automatic Identification System data. AIS data is a kind of spatio-temporal data that contains information of location data, as well as time stamps. In this way, traffic pattern is described by AIS data. Furthermore, we conduct a simulation experiment that extracts traffic pattern from the AIS data through the unsupervised technique. The proposed framework takes advantage of AIS data, which is a type of the spatiotemporal data that consists of vessel motion information, to perform the experiment. The result shows that the unsupervised framework converts useful information from raw AIS data to effective traffic pattern. The proposed method strongly supports the further research on maritime traffic pattern extraction of AIS data. Besides, an overview of the framework and the unsupervised technique for high-level maritime situation awareness is presented.

The advancement of Maritime Autonomous Surface Ships (MASS) relies heavily on data-driven knowledge discovery and situational awareness to ensure safe and efficient navigation. This PhD by Published Works comprises five peer-reviewed journal papers, contributing to the integration of Artificial Intelligence (AI), machine learning, deep learning, and Bayesian Network (BN) into key areas of maritime safety and security. The research aims to enhance risk assessment, piracy incident analysis, ship trajectory prediction, and autonomous route planning, forming a cohesive framework for intelligent decision-making in MASS operations. The first contribution focuses on data-driven risk analysis of global maritime accidents, employing a data-driven BN model to quantify accident risks, identify critical factors, and enhance safety management strategies. The second contribution addresses spatio-temporal piracy pattern analysis, utilising AI-driven methodologies to detect high-risk zones and extract patterns of piracy incidents, thereby improving security measures and navigational decision-making for MASS. The third and fourth contributions focus on ship trajectory prediction and route planning, which are crucial for autonomous navigation. A systematic review of machine learning and deep learning-based trajectory prediction methods highlights the effectiveness of data-driven models over traditional approaches. Further, an unsupervised route planning model integrates Automatic Identification System (AIS) data and machine learning techniques, enabling autonomous vessels to adapt dynamically to changing maritime conditions. The final contribution refines AIS-based trajectory prediction frameworks, leveraging machine learning methodologies to improve predictive accuracy and enhance collision avoidance capabilities. By synthesising insights from these studies, this research establishes a holistic framework for data-driven knowledge discovery and situational awareness in MASS. The integration of this research content and outcomes ensures that autonomous vessels can navigate safely, mitigate operational risks, and enhance adaptability to evolving maritime conditions. This study advances AI-driven autonomous navigation by bridging data-driven knowledge discovery with real-time maritime decision-making, providing a resilient and adaptive foundation for next-generation MASS technology.

Automatic Identification System (AIS) provides abundant near real-time information of moving vessels in the sea of the whole world and has been widely used in the fields of vessel collision avoidance, Maritime Situation Awareness (MSA) and ocean surveillance. The development of satellite-based AIS further expands the range of AIS and enables a wide converge of AIS data collection, which solves the problem of lacking AIS data in high-seas. At the same time, AIS data provides a rich source for data mining for maritime traffic analysis. In this paper, a typical clustering algorithm called K-means is applied to deal with the Space-based AIS(S-AIS) data received by “TianTuo-3” satellite developed by National University of Defense Technology. We use Elbow Rule to determine the optimal number of clusters and calculate the normalized standard deviation of COG(Course Over Ground) and SOG(Speed Over Ground) of vessels in south Africa area as their features to conduct clustering. This method is supposed to evaluate vessels' sailing stability and used in detection of low-likelihood behaviors or anomalies of vessels.

Ship Time Headway (STH) is used in maritime navigation to describe the time interval between the arrivals of two consecutive ships in the same water area. This measurement may offer a straightforward way to gauge the frequency of ship traffic and the likelihood of congestion in a particular area. STH is an important factor in understanding and managing the dynamics of ship movements in busy waterways. This paper introduces a hybrid deep learning method for predicting STH in time domain. The method integrates the Seasonal-Trend Decomposition using Loess (STL), Multi-head Self-Attention (MSA) mechanism into Long Short-Term Memory (LSTM) neural network. The STH dataset was extracted from the Automatic Identification System (AIS) through ship trajectory spatial motion, and the seasonal, trend and residual components of the decomposition were then determined from the STH dataset using the STL algorithms. MSA-LSTM is adopted to comprehensively capture the evolving patterns of STH from the sequence. Comparison studies with existing methods demonstrate the accuracy and robustness of the predictions provided by this method, indicating that the proposed method outperforms other models in terms of prediction performance and learning capabilities. By predicting STH, the method offers potential to assist maritime traffic managers and navigators in assessing ship flow, thereby enabling them to make informed decisions on navigation safety and efficiency.

Recreational boating is an increasing activity in coastal areas and its spatiotemporal overlap with key habitats of marine species pose a risk for negative noise impacts. Yet, recreational vessels are currently unaccounted for in vessel noise models using Automatic Identification System (AIS) data. Here we conduct a case study investigating noise contributions from vessels with and without AIS (non-AIS) in a shallow coastal area within the Inner Danish waters. By tracking vessels with theodolite and AIS, while recording ambient noise levels, we find that non-AIS vessels have a higher occurrence (83%) than AIS vessels, and that motorised recreational vessels can elevate third-octave band noise centred at 0.125, 2 and 16 kHz by 47–51 dB. Accordingly, these vessels dominated the soundscape in the study site due to their high numbers, high speeds and proximity to the coast. Furthermore, recreational vessels caused 49–85% of noise events potentially eliciting behavioural responses in harbour porpoises (AIS vessels caused 5–24%). We therefore conclude that AIS data would poorly predict vessel noise pollution and its impacts in this and other similar marine environments. We suggest to improve vessel noise models and impact assessments by requiring that faster and more powerful recreational vessels carry AIS-transmitters.

In order to study the ship traffic flow in LaoTieshan channel, this paper rewrites the rules of NS (Nagel-Schreckenberg) model of Cellular Automata (CA) based on AIS（Automatic Identification System） data. By referring to the idea of synchronous flow in three-phase traffic flow theory and introducing the speed adaptation mechanism into the simulation, the new model reproduces the traffic flow of LaoTieshan channel and gets the relevant fundamental diagrams. Compared with the actual flow-density diagram drawn from AIS data, it is found that the simulated flow generated by the improved model is more in line with the actual condition, and also conforms to the phase transition of traffic flow in LaoTieshan channel. The method and idea of the simulation are helpful to explain and investigate some complex situations in LaoTieshan channel.

Understanding the relationship between rosemary odor and mental workload through deep learning.

The development of China’s digital waterways has led to the extensive deployment of cameras along inland waterways. However, the limited processing and utilization of digital resources hinder the ability to provide waterway services. To address this issue, this paper introduces a novel waterway perception approach based on an intelligent navigation marker system. By integrating multiple sensors into navigation markers, the fusion of camera video data and automatic identification system (AIS) data is achieved. The proposed method of an enhanced one-stage object detection algorithm improves detection accuracy for small vessels in complex inland waterway environments, while an object-tracking algorithm ensures the stable monitoring of vessel trajectories. To mitigate AIS data latency, a trajectory prediction algorithm is employed through region-based matching methods for the precise alignment of AIS data with pixel coordinates detected in video feeds. Furthermore, an augmented reality (AR)-based traffic situational awareness framework is developed to dynamically visualize key information. Experimental results demonstrate that the proposed model significantly outperforms mainstream algorithms. It achieves exceptional robustness in detecting small targets and managing complex backgrounds, with data fusion accuracy ranging from 84.29% to 94.32% across multiple tests, thereby substantially enhancing the spatiotemporal alignment between AIS and video data.

Air pollution prediction using LSTM deep learning and metaheuristics algorithms

Wearable electroencephalography (EEG) devices for seizure detection accuracy and reliability are deep learning (DL) applications in the field of epilepsy diagnosis. In this study, we sought to increase the accuracy of seizure detection using advanced DL algorithms on the Children’s Hospital Boston - Massachusetts Institute of Technology (CHB-MIT) EEG database. First, a fully convolutional network (FCN) was trained and assessed using accuracy and recall/precision metrics, and the early stopping technique was used to avoid overfitting. To assess the performance, the FCN was evaluated in terms of various metrics, including accuracy, precision, recall, F1-score, and receiver operating characteristic (ROC)-area under the curve (AUC). In addition, two-dimensional (2D) convolutional neural networks (CNNs) and long short-term memory (LSTM) models were used to model the database, and their performance was thoroughly measured using different metrics, graphs, and confusion matrices. Using LSTM variants, such as standard LSTM, bidirectional LSTM, stacked LSTM, and LSTM attention mechanisms, hybrid convolutional LSTM (ConvLSTM) models were trained and compared. The comparison was conducted based on the training and validation accuracy and loss, as well as the graphs resulting from the precision–recall curves. Apart from DL approaches, EEG signal analysis using time–frequency techniques, such as wavelet transform and short-time Fourier transform, has also been investigated. These methods assisted in the analysis of the time–frequency features of EEG signals in combination with DL models. This study demonstrates that the performance of wearable EEG devices can be augmented using a combination of DL and seizure signal processing techniques. The FCN achieved an accuracy of 92%, with a recall for seizures of 33%, an F1-score of 0.03, and strong ROC-AUC results. The 2D CNN achieved 96% accuracy, a seizure recall of 70%, an F1-score of 0.12, and an ROC-AUC score of 78%. The baseline LSTM struggled with effectiveness at 53% accuracy with a seizure recall of 18%. In contrast, the LSTM model, which incorporated synthetic minority oversampling technique (SMOTE) balancing, was able to reach improvements of up to 89% accuracy, with a precision of 91%, a recall of 86%, an F1-score of 0.89, and a strong ROC curve performance. Among the models, the LSTM with SMOTE was the best performer, with 89% accuracy, 91% precision, 86% recall, and an F1-score of 0.89. These results provide evidence that applying techniques for data balancing in combination with certain DL network architectures significantly improves the detection of seizures using wearable EEG devices worn on the body. We believe that real-time monitoring and high-performance systems are feasible using optimized DL frameworks. The analysis of the performance of different models allows for the understanding of the possibilities of optimizing the architectures of DL algorithms for the modern diagnosis of epilepsy in real time. The source code used to carry out the experiments is publicly available at CHB-MIT EEG Dataset Python Scripts (https://www.kaggle.com/code/adnankust/adnaneeg1).

Automatic Identification System (AIS) allows the ships to broadcast important kinematic and static information, and is one of the most commonly used tools for ship traffic monitoring as individual tracks can be inferred from chronological sequence of the ships' AIS messages. In this paper, we propose two deep learning based methods for AIS data association. The first method predicts a ship's position at the time of a new message and then computes the association probability. The second method computes the association probability directly without ship position interpolation. Both methods use only three AIS data attributes for inference: longitude, latitude and time. We validate the proposed methods' performance with real AIS dataset and show that they achieve reliable data association.

Assessing the fishing footprint using data integrated from different tracking devices: Issues and opportunities