• Your email
    Your password
  •  Lang.
      • Artificial Intelligence takes root in the maritime sector

      • In June, the French Tech Brest+, supported by the Technopôle Brest-Iroise, organised the second edition of AI Days, two days of conferences and workshops dedicated to Artificial Intelligence. This was an opportunity for theCampus mondial de la mer to host a round table on the challenges of AI in the maritime sector and to highlight a few initiatives in this field.
      • Round table on AI in the maritime sector during AI Days 2022 with Alice de Joux (Technopôle Brest-Iroise), Simon Benaiïchouche (eOdyn, IMT Atlantique), John Queffélec (BluePulse) and Christophe Desbois (Blue HUB from ISblue).

        A broad spectrum of AI applications in the maritime sector

        In the maritime field, AI is everywhere. The nautical sector and in particular that of ocean racing have already embarked on AI, as highlighted by the round table moderated by Rachel Portal-Sellin (Finistère Mer Vent) during the AI Days 2021. The use of SenX's AI solutions by AIM45 for the analysis and visualisation of sensor data dispatched on racing boats is an example.

        President of the European Commision Ursula von der Leyen at the One Ocean Summit, Brest 2022.  ©European Commission

        AI is also being mobilised for ocean observation and modelling, with European Commission President Ursula van der Leyen's push at the One Ocean Summit to create a digital twin of the ocean by 2024. By combining next-generation ocean modelling and other scientific, economic and statistical data, this platform will strengthen the necessary knowledge sharing for a use of marine resources compatible with the preservation of marine ecosystems.

        These issues will be addressed again in September during the "AI in Meteorology and Oceanography" session organised during Sea Tech Week® by IMT Atlantique and the Japanese Institute RIKEN , registration is open!

        The observation of underwater biodiversity is increasingly integrating AI methods. The AI 4 KOSMOS project of Konk Ar Lab in Concarneau, with the support of Ifremer, uses machine learning techniques to automate fish recognition and counting tasks. This project is being expanded within Blue HUB with a POC manufactured at La Forge, ENIB's FabLab, which will be deployed for sea trials within the Iroise Marine Nature Park in 2023.

        For the study of benthic communities, the emblematic EUR ISBlue Blue Revolution project brings together Ifremer, ENIB, Lab-STICC and other international academic partners to develop microscopy methods, in situ imaging and AI-based classification tools to generate quantitative and functional data. Based on these bioindicator species, the aim is also to assess global environmental change.

        Sensors dedicated to observation are therefore being fitted to instruments from the deep sea to space. Also booming for their exploration capabilities, drones are at the heart of dual innovations, with, for example, the ECA group implementing Artificial Intelligence methodologies applied to robotic and autonomous systems.

        More broadly, the defence sector is mobilising AI techniques, whether for mine detection by Thales, or when the Ecole Navale is developing real-time decision support applications and analysing AIS (Automatic Identification System) positioning data to detect suspicious behaviour.

        In the maritime transport sector, these applications are just as wide-ranging and were illustrated by the speakers at the round table.

        AI is revolutionising maritime transport

        AI is being used to optimise maritime routing, to make navigation safer. When we talk about the ships of the future, we are referring to smarter, more autonomous and more energy-efficient ships. When we talk about the ports of the future, we are referring to optimised management of flows, stocks and maintenance.

        The company eOdyn reconstructs marine surface currents from AIS signals and thus offers SeaWaze, a maritime routing optimisation solution that allows for significant energy savings. Simon Benaïchouche demonstrated during his thesis, as part of joint work between eOdyn and IMT Atlantique, that deep learning allows a qualitative leap in the generation of current fields.

        BluePulse, a company based in Aix-en-Provence and Brest, is developing AI solutions for refrigerated maritime transport, a fast-growing sector with a high environmental impact. Using patented Deep Learning and Supervised Learning models, the solutions developed by BluePulse exploit the IoT data increasingly present in the maritime sector, and reduce this impact by developing predictive maintenance programmes for refrigerated containers, solutions for optimising the routing of container doors and detecting illegal intrusion into containers using satellite images.

        An ecosystem favourable to the development of AI

        Such innovations in the field of AI are based on excellent research and frameworks conducive to research-business interactions, such as the AI OceaniX chair headed by Ronan Fablet, professor at IMT Atlantique. This research and teaching chair on the theme of ocean dynamics brings together research laboratories, institutional, industrial (including eOdyn) and international partners. Together, these players are taking up the challenge of improving the capacity to predict and simulate ocean parameters, as well as optimising observation resources.

        The IA 4 KOSMOS team, Ocean Hackathon® 2021 at Brest

        Open innovation encourages the emergence of projects from start-ups and innovative companies. For the past 7 years, Ocean Hackathon® has enabled companies to boost their projects by relying on multidisciplinary teams and by having access to maritime data. In 2021, some of the teams from Brest were able to benefit from the support of coaches specialised in AI, including the AI 4 KOSMOS project mentioned above.

        The issue of digital infrastructures is central to the management and processing of the data flow. Within the framework of the Blue HUB and the OceaniX AI Chair, groups of students can access the DATARMOR supercomputer and its "Sandbox" to test algorithms on new data sets.

        Local companies can also count on the adaptation of the training offer and thus recruit students trained in AI. IMT Atlantique, ISEN and ENSTA Bretagne have created modules dedicated to AI in their engineering courses, with possible specialisations in virtual reality at ENIB, or in supply chain applied to ports at Brest Business School. Within the framework of ISblue, the Ocean Data Science (SDO) course of the international Marine Sciences master's degree, co-accredited by IUEM at UBO, ENSTA Bretagne and IMT Atlantique, creates the interfaces between AI and physical modelling of the ocean. This excellent training offer is completed by the opening of the Microsoft by Simplon AI school at ISEN two years ago, offering people who are far from employment the opportunity to specialise in AI and to join companies in the ecosystem on a sandwich course.

        Research, training and innovation, the winning triptych for the development of AI in Western Brittany!

      • Published on 08/31/2022
Campus mondial de la mer
Technopôle Brest-Iroise
525, Avenue Alexis de Rochon
29280 Plouzané
Contact us
  • Brest Métropole
  • Région Bretagne
  • https://www.tech-brest-iroise.fr/
Newsletters: Unsubscribe | Mailings: Unsubscribe
Legals | Sitemap | Powered by diasite
Designed by diateam