Making MAST fusion tokamak data open and FAIR

Go back to programme

Making MAST fusion tokamak data open and FAIR

  • Watch

    * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

    • Watch

      Register

      * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

    The fusion community has historically operated within silos. Sharing data, when it does occur, tends to be ad-hoc, in non-standard formats, with little resources available for carefully curated provenance descriptions. For an open fusion database, we as a community need to develop a shared ontology for describing our data, resulting in interoperable datasets, which can enable a new generation of fusion researchers to apply modern AI/ML methods on larger sets of data than ever before to deepen our understanding of magnetic confinement fusion (MCF) and tokamak operation.

     

    In recent years, the fusion community has begun to take this challenge seriously. From the development of the IMAS (ITER Integrated Modelling and Analysis Suite) Data Dictionary, intended to be applicable to fusion data from a wide range of devices, to more strategic assessments of the current landscape and demonstrator technology from the FAIR4fusion project, and now with the IAEA CRP on AI for Fusion, the need for a better way of accessing and working with data to accelerate fusion R&D is firmly on the agenda.

     

    The UK Atomic Energy Authority (UKAEA) intends to lead this effort by developing an open data access platform for data from their Mega-Ampere Spherical Tokamak (MAST) that conforms to FAIR principles (Findable, Accessible, Interoperable, Reusable), as an exemplar for other fusion experiments, and alongside the objectives of the IAEA AI for Fusion coordinated research project. Some use cases that such a database could enable are outlined, along with a discussion about the various decisions and challenges that need to be overcome as this effort progresses. This includes deciding on how to adapt to a common standard for data, where to store the data, how to update it, and crucially, how modern technology stacks will enable scalable, performant access to data.

     

    A number of examples in other scientific fields will be discussed, outlining their merits and limitations with regard to the previously mentioned use cases, and how this is informing the development of an open and FAIR fusion database for MAST data.

     

    This live event includes a 30-minute networking event hosted on the AI for Good Neural Network. This is your opportunity to ask questions, interact with the panelists and participants and build connections with the AI for Good community.

    Share this session
    • Watch

      Register

      * Register (or log in) to the AI4G Neural Network to add this session to your agenda or watch the replay

    • Start date
      16 May 2023 at 17:00 CEST Geneva | 11:00-12:30 EDT, New York | 23:00-00:30 CST, Beijing
    • End date
      16 May 2023 at 18:30 CEST Geneva | 11:00-12:30 EDT, New York | 23:00-00:30 CST, Beijing
    • Duration
      90 minutes (including 30 minutes networking)
    • Programme stream
    • UN SDGs
    In partnership with

    Are you sure you want to remove this speaker?