Data-driven methods open up unprecedented possibilities for maritime surveillance using Automatic Identification System (AIS) data. In this work, we explore deep learning strategies historical AIS observations to address the problem of predicting future vessel trajectories with a prediction horizon several hours. We propose novel sequence-to-sequence trajectory models based on encoder-decoder recurrent neural networks (RNNs) that are trained data predict samples given previous observations....

Abstract To prevent the outbreak of Coronavirus disease (COVID-19), many countries around world went into lockdown and imposed unprecedented containment measures. These restrictions progressively produced changes to social behavior global mobility patterns, evidently disrupting economic activities. Here, using maritime traffic data collected via a network Automatic Identification System (AIS) receivers, we analyze effects that COVID-19 pandemic measures had on shipping industry, which...

We present a novel method for predicting long-term target states based on mean-reverting stochastic processes. use the Ornstein-Uhlenbeck (OU) process, leading to revised state equation and time scaling law related uncertainty that in long term is shown be orders of magnitude lower than under nearly constant velocity (NCV) assumption. In support proposed model, an analysis significant portion real-world maritime traffic provided.

We propose an unsupervised procedure to automatically extract a graph-based model of commercial maritime traffic routes from historical Automatic Identification System (AIS) data. In the proposed representation, main elements patterns, such as maneuvering regions and sea-lanes, are represented, respectively, with graph vertices edges. Vessel motion dynamics defined by multiple Ornstein-Uhlenbeck processes different long-run mean parameters, which in our approach can be estimated change...

In this paper, we address the problem of predicting vessel trajectories based on Automatic Identification System (AIS) data. The goal is to learn predictive distribution maritime traffic patterns using historical data during training phase, in order be able forecast future target trajectory samples online basis both extracted knowledge and available observation sequence. We explore neural sequence-to-sequence models Long Short-Term Memory (LSTM) encoder-decoder architecture effectively...

Maritime surveillance (MS) is crucial for search and rescue operations, fishery monitoring, pollution control, law enforcement, migration national security policies. Since the early days of seafaring, MS has been a critical task providing in human coexistence. Several generations sensors detailed maritime information have become available large offshore areas real time: radar 1950s automatic identification system (AIS) 1990s among them. However, ground-based radars AIS data do not always...

A novel anomaly detection procedure based on the Ornstein–Uhlenbeck (OU) mean-reverting stochastic process is presented. The considered a vessel that deviates from planned route, changing its nominal velocity <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\boldsymbol{v}_0$</tex-math></inline-formula> . In order to hide this behavior, switches off automatic identification system (AIS) device for time...

Since the beginning of 2020, outbreak a new strain Coronavirus has caused hundreds thousands deaths and put under heavy pressure world's most advanced healthcare systems. In order to slow down spread disease, known as COVID-19, reduce stress on structures intensive care units, many governments have taken drastic unprecedented measures, such closure schools, shops entire industries, enforced social distancing regulations, including local national lockdowns. To effectively address pandemics in...

Recent deep learning methods for vessel trajectory prediction are able to learn complex maritime patterns from historical automatic identification system (AIS) data and accurately predict sequences of future positions with a horizon several hours. However, in surveillance applications, reliably quantifying the uncertainty can be as important obtaining high accuracy. This article extends frameworks tasks by exploring how recurrent encoder–decoder neural networks tasked not only but also yield...

Today, the maritime domain is at cusp of a new era, driven by technological advances in automation, robotics, multisensor perception, and artificial intelligence (AI), together with digitalization connectivity. Smart ship infrastructure technology, remotely controlled autonomous operation to improve safety, security, cost efficiency, sustainability are future transportation <xref ref-type="bibr" rid="ref1" xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Maritime surveillance (MS) is of paramount importance for search and rescue operations, fishery monitoring, pollution control, law enforcement, migration national security policies. Since ground-based radars automatic identification system (AIS) do not always provide a comprehensive seamless coverage the entire maritime domain, use space-based sensors crucial to complement them. We reviewed technologies MS in first part this work, titled "Space-based Global Surveillance. Part I: Satellite...

In this paper we present how automatic maritime anomaly detection tools can be successfully applied in real-world situations such as the major event of container vessel <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Ever Given</i> , which grounded Suez Canal on March 23rd 2021. The detector is designed to process available sequence Automatic Identification System (AIS) reports, information from ground-based or satellite radar systems if...

Abstract To prevent the outbreak of Coronavirus disease (COVID-19), numerous countries around world went into lockdown and imposed unprecedented containment measures. These restrictions progressively produced changes to social behavior global mobility patterns, evidently disrupting economic activities. Here, using maritime traffic data, collected via a network Automatic Identification System (AIS) receivers, we analyze effects that COVID-19 pandemic measures had on shipping industry, which...

The underwater environment poses numerous challenges and risks, making Unmanned Underwater Vehicles (UUVs) an indispensable alternative to human operators. Numerous Remotely Operated (ROVs) Autonomous (AUVs) have been developed as a valuable resource in broad spectrum of operations. However, the deployment operation UUVs face significant due unique that critically affects Positioning, Navigation, Timing (PNT) performance, it incomparable above-water applications. This discrepancy...

In this letter, we study the problem of estimating long-run mean Ornstein-Uhlenbeck (OU) stochastic process and its effect on long-term prediction future vessel states, which is a crucial for Maritime Situational Awareness (MSA). We employ sample estimator (SME) to estimate key OU parameter from observations, computing closedform SME covariance error in both random constant sampling time regimes, providing fundamental building block overall state covariance. show also that is: √n-consistent...

A novel anomaly detection procedure is presented, based on the Ornstein-Uhlenbeck (OU) mean-reverting stochastic process. The considered a vessel that deviates from planned route, changing its nominal velocity. In order to hide this behavior, switches off Automatic Identification System (AIS) device for certain time, and then tries revert previous decision has be taken either declaring deviation happened or not, relying only upon two consecutive AIS contacts. proper statistical hypothesis...

Seaports play a vital role in the global economy, as they operate connection corridors to all other modes of transport and engines growth for wider region. But ports today are faced with numerous unique challenges them remain competitive, significant investments required. In support greater transparency policy making, decisions regarding investment need be supported by data-driven intelligence. It is often an overlooked fact that seaports do not static over time; such spatial units evolve...

In this work, we propose a data driven trajectory forecasting algorithm that utilizes both recorded historical and streaming observations. The performs Bayesian inference on directed graph the walks which represent stochastic change point models of classes. Parameter distributions these are learnt from trajectories. Forecasting is then made by calculating class - or, walk- probabilities corresponding predictive for given stream location velocity This approach tailored maritime domain...

This paper presents an unsupervised approach to extract maritime Patterns of Life (PoL) from historical Automatic Identification System (AIS) data based on a low-dimensional synthetic representation ship routes. Recent advances in long-term vessel motion modeling through Ornstein-Uhlenbeck mean-reverting stochastic processes allow encode knowledge about traffic via compact graph-based model where waypoints are graph vertices and the connections between them, i.e., navigational legs, edges....

The explosions on 26 September 2022, which damaged the Nord Stream gas pipelines, have highlighted need and urgency of improving resilience critical undersea infrastructures (CUIs). Comprising pipelines power communication cables, these connect countries worldwide are for global economy stability. An attack targeting multiple such could potentially cause significant damage greatly affect various aspects daily life. Due to increasing number CUIs, existing underwater surveillance solutions, as...

Ship traffic monitoring is a foundation for many maritime security domains, and system specifications underscore the necessity to track vessels beyond territorial waters. However, in open seas are seldom continuously observed. Thus, problem of long-term vessel prediction becomes crucial. This paper focuses attention on performance assessment Ornstein-Uhlenbeck (OU) model prediction, compared with usual well-established nearly constant velocity (NCV) model. Heterogeneous data, such as...

Piecewise mean-reverting stochastic processes have been recently proposed and validated as an effective model for long-term object prediction. In this paper, we exploit the Ornstein-Uhlenbeck (OU) dynamic to represent anomaly any deviation of long-run mean velocity from nominal condition. This amounts modeling unknown switching control input that can affect dynamics object. Under model, problem joint detection tracking be addressed within Bayesian random set framework by means a hybrid...

In this article, we present a fully analytical model for the evaluation of electromagnetic (EM) field scattered from composite target in generic bistatic configuration. The scenario comprises rectangular parallelepiped with smooth dielectric faces lying over rough background surface, modeled as stochastic process. single- and multiple-bounce scattering contributions arising target, background, their interactions have been derived under Kirchhoff approximation (KA)-geometrical optics (GO)...

During the course of an epidemic, one most challenging tasks for authorities is to decide what kind restrictive measures introduce and when these should be enforced. In order take informed decisions in a fully rational manner, onset critical regime, characterized by exponential growth contagion, must identified as quickly possible. Providing rigorous quantitative tools detect such represents important contribution from scientific community proactively support political decision makers. this...