Machine learning methods are increasingly being used in the maritime domain to predict traffic anomalies and to mitigate risk, for example avoiding collision and groundingaccidents. However, most machine learning systems used for detecting such issues hasbeen trained predominately on single data sources such as vessel positioning data. Hence,it is desirable to support the means to combine different sources of data - in the trainingphase - to allow more complex models to be built. In this thesis, we propose a multi-data pipeline for accumulating, decoding, preprocessing, and merging Automatic Identification System (AIS) data with weather datato train time series based deep learning models. The pipeline comprises several REST APIsto connect and listen to the data sources, and storing and merging them using StructuredQuery Language (SQL). Specifically, the training pipeline consists of an AIS NMEA message decoder, weather data receiver, and a Postgres database for merging and storing thedata sources. Moreover, the pipeline was assessed by training a TensorFlow vRNN model.The proposed pipeline approach allows flexibility in the inclusion of new data sources toeffectively build models for the maritime domain as well as other traffic domains that usespositioning data.

The surveillance of large sea areas typically involves  the analysis of huge quantities of heterogeneous data.  In order to support the operator while monitoring maritime traffic, the identification of anomalous behavior or situations that might need further investigation may reduce operators' cognitive load. While it is worth acknowledging that existing mining applications support the identification of anomalies, autonomous anomaly detection systems are rarely used for maritime surveillance. Anomaly detection is normally a complex task that can hardly be solved by using purely visual or purely computational methods. This thesis suggests and investigates the adoption of visual analytics principles to support the detection of anomalous vessel behavior in maritime traffic data. This adoption involves studying the analytical reasoning process that needs to be supported,  using combined automatic and visualization approaches to support such process, and evaluating such integration. The analysis of data gathered during interviews and participant observations at various maritime control centers and the inspection of video recordings of real anomalous incidents lead to a characterization of the analytical reasoning process that operators go through when monitoring traffic. These results are complemented with a literature review of anomaly detection techniques applied to sea traffic. A particular statistical-based technique is implemented, tested, and embedded in a proof-of-concept prototype that allows user involvement in the detection process. The quantitative evaluation carried out by employing the prototype reveals that participants who used the visualization of normal behavioral models outperformed the group without aid. The qualitative assessment shows that  domain experts are positive towards providing automatic support and the visualization of normal behavioral models, since these aids may reduce reaction time, as well as increase trust and comprehensibility in the system. Based on the lessons learned, this thesis provides recommendations for designers and developers of maritime control and anomaly detection systems, as well as guidelines for carrying out evaluations of visual analytics environments. / Maria Riveiro is also affiliated to Informatics Research Centre, Högskolan i Skövde / Information Fusion Research Program, Högskolan i Skövde

Orientadores: Luiz Leduino de Salles Neto, Antonio Carlos Moretti / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Matemática, Estatística e Computação Científica / Made available in DSpace on 2018-08-23T22:02:01Z (GMT). No. of bitstreams: 1 Cereser_BrunoLuisHonigmann_M.pdf: 8696662 bytes, checksum: b4ca681ecf08f10afa47d00b3a879e9e (MD5) Previous issue date: 2013 / Resumo: O transporte marítimo é fundamental para grandes empresas de caráter global. Em 2009 80% do comércio mundial foi realizado via transporte marítimo e desde então esse número só tem crescido. Devido a esse alto fluxo nos portos mundiais, surgiu o problema de alocação de navios em berços, que tem por finalidade estabelecer uma sequência de atendimento de navios para cada berço disponível no cais, visando minimizar o tempo total da operação e espera dos navios dentro dos portos. Nesse trabalho o problema de alocação de navios em berços será abordado considerando berços dependentes, e também a divisão das máquinas portuárias em berços consecutivos aplicados a terminais do Porto de Tubarão (ES). Para a representação matemática dos processos desses terminais do porto de tubarão foram desenvolvidos novos modelos matemáticos. Os testes computacionais foram exitosos / Abstract: The maritime transportation is essential to global companies. In 2009 80% of global trade was made by maritime transportation, and over the year this number has raised. Because of this high flux, the BAP (Berth Allocation Problem) arose. The BAP have intent to discovery the vessel allocation sequence in every berth with the minimum total wait time. In this work, the berths of BAP will be considered dependents, we also considered the division of portuary machines between consecutive berths. The objective of this work is the mathematical representation of the terminals Produtos Diversos and Praia Mole of the Tubarão port. For this reasons we propose here unprecedented mathematical models / Mestrado / Matematica Aplicada / Mestre em Matemática Aplicada

Förbränning av traditionellt fartygsbränsle medför utsläpp av ämnen som har negativa effekter på miljön. Uppkomsten av strängare internationella regler gällande utsläpp från sjöfartstrafiken kräver ett skifte från det konventionella marina bränslet tjockolja till alternativ med lägre utsläpp. Vågkraft och metanol är två möjliga alternativ inom sjöfarten som kan uppfylla de internationella regleringarna för utsläpp, och har i denna studie jämförts med tjockolja. Beräkningar av emissioner, kostnad och den möjliga producerade effekten utfördes med hjälp av referensfartyget M/V Fidelio som underlag. Studien visar att metanol och vågkraft bidrar med signifikant lägre utsläpp, vilket tyder på stora möjligheter att uppfylla de uppsatta internationella utsläppskraven. Däremot uppfyller de inte referensfartygets effektbehov och är kostsamma i relation till deras möjliga effektuttag. För att minska fartygens miljöpåverkan, lämpar det sig i dagsläget bättre att använda metanol och vågkraft som komplement till andra drivmedel. / Combustion of traditional ship fuels causes emissions of substances that have negative impacts on the environment. Adoption of stricter international regulations regarding emissions from the maritime traffic requires possible substitution of the widely used conventional heavy fluid oil, with lower emissions. Wave power and methanol are two possible substitutions in the maritime traffic that could fulfill the adopted international regulations regarding emissions, and have in this study been compared to heavy fluid oil. Calculations regarding emissions, costs and the possible power generation were performed using the ship M/V Fidelio as a basis. The study presents that wave power and methanol have significant lower emissions, which increases their opportunity to fulfill the international emission regulations. However, these options fail to fulfill the ship’s power demand. Furtheremore, these alternative solutions are costly in relation to their power generation. Currently it is more suitable to use methanol and wave power as a supplement to other fuels, in order to reduce the maritime traffic’s impact on the environment.

This qualitative study has evaluated the economic development progression in the Russian Arctic over the 2008-2020 time-period. Departing from the qualitatively formulated official objectives in the Russian Federation’s Policy for the Arctic to 2020 (signed in 2008), this investigation sought to trace the implementation process and analyse the extent in which they have been met. A geoeconomic theoretical framework was used to guide this study through a dual case-centric process tracing and policy evaluation methodology. The data in this investigation mainly originated from news sources, think tanks, Arctic-specific foundations, and official reports from the Arctic Council and the Nordic Council of Ministers. The investigation was not hypothesis driven but rather attempted to assess the extent to which Russia’s stated objectives corresponded with the economic reality in the Russian Arctic region. The investigation has focused on three separate but interrelated elements of the regional economic development, namely maritime activity, hydrocarbon resource development and demographic trajectory. The analysis found that economic activity in the Russian Arctic regions has increased over the time-period and the stated policy objectives are to be considered fulfilled.

Le trafic maritime est le principal contributeur des bruits sous-marins anthropique : depuis les années 1970, l’augmentation du trafic maritime hauturier a provoqué dans certaines zones une augmentation du bruit ambiant de plus de 10 dB. En réponse à cette préoccupation, la Directive Cadre pour la Stratégie pour le Milieu Marin (DCSMM) recommande un suivi acoustique. Peu d’études s’intéressent à l’activité côtière et aux bruits rayonnés par les petites embarcations ainsi qu’à leurs conséquences sur la faune marine alors que ces environnements côtiers sont les pourvoyeurs de 41.7 % des services écosystémiques produits par les océans.A mi-chemin entre le monde académique et le monde industriel, le travail présenté aux différents questions scientifiques et industrielles sur la thématique du trafic côtier, en termes de l’étude de son influence dans le paysage acoustique et de capacité à détecter et classifier les embarcations côtières.En l’absence d’information sur le trafic maritime côtier, un protocole d’identification visuelle par traitement d’images GoPro® produisant les mêmes données que l’AIS (position, vitesse, taille et type d’embarcation) est proposé et permet la création de carte du trafic maritime sur un disque de 1.6km de rayon. D’un point de vue acoustique, le trafic est caractérisé par deux descripteurs acoustiques, le SPL lié à la distance du bateau le plus proche et l’ANL caractérisant le nombre de bateaux dans un disque de 500 m de rayon. Le suivi spatio-temporel de ces descripteurs permet d’identifier l’impact du trafic maritime dans le paysage acoustique des environnements côtiers. La détection et la classification sont réalisées après caractérisation individuelle du bruit par un ensemble de paramètres acoustiques et par l’utilisation d’algorithmes d’apprentissage supervisé. Un protocole spécifique pour la création de l’arborescence de classification est proposé par comparaison des données acoustiques aux caractéristiques physiques et contextuelle de chaque bateau.Les travaux présentés sont illustrés sur la flottille d’embarcations côtières présente dans la baie de Calvi (Corse) durant la saison estivale. / Marine traffic is the main contributor to anthropogenic underwater noise: since the 1970s, the increase in deep-sea shipping has increased the ambient noise by more than 10 dB in some areas. In response to this concern, the Marine Strategy Framework Directive (MSFD) recommends acoustic monitoring. Few studies are concerned with coastal activity and the noises radiated by small craft while these coastal environments are the purveyors of 41.7% of the ecosystem services produced by the oceans.Between the academic and the industrial world, this PhD was to answer the different scientific and industrial questions on the topic of the coastal traffic in terms of the influence in the soundscape and the detection and classification of the coastal craft.Without information on the coastal maritime traffic, a visual identification protocol is proposed using GoPro® images processing and produced the same data as the AIS (position, speed, size and type of craft); It allows to create maritime traffic maps on a disk of 1.6km radius. The traffic is characterized by two acoustic descriptors: the SPL linked to the distance of the nearest boat and the ANL linked to the number of boats present in a 500 m radius disc. The spatiotemporal monitoring of these descriptors allows to identify the impact on the maritime traffic on the coastal acoustic landscape. The acoustic detection and the classification are performed after individual characterization of the noise by a set of acoustic parameters and using of supervised machine learning algorithm. A specific protocol for the creation of the classification tree is proposed by comparing the acoustic data with the physical and contextual characteristics of each boat.The methods are applied on the flotilla of coastal boats present in the Bay of Calvi (Corsica) during summer.

As Arctic warms twice as fast as the rest of the world and the polar ice-cap melts, the strategic importance and geopolitical significance of the Bering Strait as the only maritime gateway between the world's fastest-developing and dynamic regions, the Asia Pacific and the Arctic region, will steadily grow. The climate change triggered the reduction of the Arctic ice-cap, which results in increased maritime traffic activity as new shipping routes are becoming more viable and mineral resources more accessible. This has a tremendous impact on the region as it opens it to economic development but at the same time it puts strain on its fragile environment. The goal of the paper is to affirm that the geopolitical significance of the Bering Strait is increasing and will increase in the near future. The work examines the causes of the Bering Strait region's rise, its characteristics, challenges and opportunities together with the assessment of major regional actors' interests and approaches towards the region. Next, paper focuses The emphasis is put especially on the role of the United States and on their attitude toward the Bering Strait region. It therefore pays attention namely on the US policy and goals in the region, and on the implications the Strait's growing global importance has for the United...

L’étude de la pollution troposphérique inclut l’étude des gaz traces provenant de sources anthropiquesdiverses, dont l’impact varie de l’échelle locale à globale. Pour caractériser cette pollution, il est nécessairede mesurer avec précision les concentrations en polluants. Dans le cadre de projets européens, troiscampagnes aéroportées ont été conduites, dans le bassin méditerranéen, en Afrique de l’ouest et enMalaisie, pour mesurer différents types de polluants grâce, entre autres, à un spectromètre à lasersinfrarouges, SPIRIT, capable de mesurer rapidement en ligne de faibles variations de NO2 et CO. Les jeuxde données de ces gaz traces (et d’autres) ont été combinés avec un modèle lagrangien de dispersion departicules, FLEXPART, pour identifier différentes sources de pollution locales et régionales de l’air. Cettethèse présente ainsi les mesures et le travail de modélisation entrepris afin de définir les sources depollution de chaque région. Il a été montré que le bassin méditerranéen a été impacté, de la moyenne à lahaute troposphère pendant l’été, par des émissions de feux de biomasse venant de Sibérie et du continentnord-américain. Les régions ouest-africaine et malaisienne sont aussi impactées par des émissions de feuxde biomasse transportées depuis l’Afrique centrale dans le premier cas, et de feux locaux dans le second.En outre, ces régions sont influencées par des émissions provenant de l’exploitation du pétrole et du traficmaritime. FLEXPART a été utilisé afin d’identifier l’origine des pics de pollution mesurés au cours descampagnes. Nous avons ainsi montré que les conditions atmosphériques, qui définissent la hauteurd’injection du panache de la plateforme pétrolière, ainsi que le flux d’émission sont des paramètres clésdans la caractérisation des mesures par le modèle. / The study of tropospheric pollution includes the study of trace gases coming from various anthropogenicsources that can impact scales ranging from local to global. To characterise this pollution, it is necessary tobe able to measure with precision pollutant concentrations. Within the frame of European projects, threeairborne campaigns in the Mediterranean Basin, in West Africa and in Malaysia were conducted to measuredifferent types of pollutants thanks to, among others, an infrared laser spectrometer, SPIRIT, able to rapidlymeasure on-line small variations in NO2 and CO. The data sets of these trace gases (and others) have beencombined with a Lagrangian model of particle dispersion, FLEXPART, to fingerprint different sources of localand regional air pollution. Thus, this thesis presents the measurements and the modelling work undertakenin order to define the sources of pollution of each region. It has been shown that the Mediterranean Basinwas impacted, in the mid to upper troposphere during summer, by biomass burning emissions coming fromSiberia and the Northern American continent. West African and Malaysian regions are also impacted bybiomass burning emissions transported from central Africa in the first case, and from local fires in thesecond. In addition, those regions are influenced by emissions coming from oil exploitation and maritimetraffic. FLEXPART was used in order to identify the origin of the pollution peaks measured during thecampaigns. It has been shown that atmospheric conditions, which define the injection height of the oilplatform plume, and also the emission flux are key parameters in the characterisation of the measurementsby the model.