Exploiting the Potential of the Future “Maritime Big Data”

Abstract : Today, most of the operational “abnormal behaviour” detection algorithms primarily operate kinematic rules on the vessel tracks provided by the AIS. To be more effective, they must be associated to additional “context data”, however insuring this data access is today a challenge, due to the disparity of registers and the fragmentation of actors… For example, detecting suspicious “associations of ships” requires already a complex data mining to detect indirect common ownership through the myriad of cascaded legal entities used for formal ownership and registration; as another example, maritime security actors know the importance of environmental factors (sea state, fog, clouds, moon etc) when assessing the risk of illegal passages or piracy attacks at night – but gathering the right local weather forecast data in association with abnormal behaviour detection algorithms for maritime radar systems is also a challenge. While many more data and metadata should be browsed from all the existing maritime reporting systems and data repositories to cross-check systematically the declared versus actual behaviour of commercial, fishing and leisure ships, we shall start developing a new data access thinking to benefit from the progressive deployment of the EU-wide CISE which is approaching its pre-operational validation milestone and should be largely mature by 2020, materializing a break-through in terms of data access for the maritime security communities. In parallel, another key enabler is the capacity to collect the “local picture” gathered by genuinely cooperating shipping (sightings and nav radar) with the “big picture” (VMS, AIS, S-AIS, LRIT, satellite imaging…). The VDES will provide a very effective data uplink as an alternative to broadband maritime SatComs at the same 2020 horizon. Other planned technological gap-fillers deserve to be integrated in future data processing strategies: new space projects aim at solving the data synchronicity challenge by co-locating SAR, S-AIS and VDES payloads for specialized maritime surveillance constellations; EDRS allows downloading LEO maritime surveillance data streams in near real-time from anywhere on Earth; smart and fast embarked data processing will downsize the “rising tide” of data, extracting and tagging straight away the mere fraction of data requiring prompt human attention… This paper will aim at delivering a sort of “wake-up call” to integrate this new data access paradigm in the current research on maritime knowledge discovery associated to the detection of safety and security threats: 2020 is tomorrow, we shall think, develop and test our toolbox at the whole scale of this “Big Data”.
Type de document :
Communication dans un congrès
Liste complète des métadonnées

Littérature citée [8 références]  Voir  Masquer  Télécharger

https://hal-mines-paristech.archives-ouvertes.fr/hal-01421611
Contributeur : Magalie Prudon <>
Soumis le : jeudi 22 décembre 2016 - 15:49:20
Dernière modification le : lundi 12 novembre 2018 - 10:53:22
Document(s) archivé(s) le : lundi 20 mars 2017 - 23:19:20

Fichier

Garnier_Napoli_MKDAD2016.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01421611, version 1

Citation

Bernard Garnier, Aldo Napoli. Exploiting the Potential of the Future “Maritime Big Data”. Maritime Knowledge Discovery and Anomaly Detection Workshop, Jul 2016, Ispra, Italy. pp.24-27 - ISBN 978-92-79-61301-2. ⟨hal-01421611⟩

Partager

Métriques

Consultations de la notice

186

Téléchargements de fichiers

219