BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260724T151408Z
LOCATION:Bldg. 6 - 003
DTSTART;TZID=Europe/Stockholm:20260629T133000
DTEND;TZID=Europe/Stockholm:20260629T153000
UID:submissions.pasc-conference.org_PASC26_sess103@linklings.com
SUMMARY:MS1B - Applications of AI and ML Towards Addressing Magnetic Fusio
 n Challenges
DESCRIPTION:Access the recording\n\nOrganizer(s): Stephan Brunner (Swiss P
 lasma Center, EPFL), Eric Sonnendruecker, and Florian Hindenlang (Max Plan
 ck Institute for Plasma Physics, Garching)\n\nNuclear fusion is increasing
 ly seen as a credible complement to renewable energy, driven today by both
  public and growing private investments. The leading fusion concepts, toka
 maks and stellarators, rely on magnetic confinement. Fusion plasmas exhibi
 t multiscale electromagnetic instabilities, from machine-scale disruptions
  to microscopic turbulence. The development of these systems thus requires
  extensive high-performance computing to understand their complex physics 
 and to address the engineering challenges. Fully kinetic models are prohib
 itively expensive, so hierarchies of reduced models are needed. Machine le
 arning has therefore become a powerful tool, enabling for example fast sur
 rogate models for plasma equilibrium solvers and transport modules. Such m
 odels are also developed for building sub-system modules integrated into c
 omprehensive physics-engineering digital twin environments. In some cases,
  optimization is such that these models can be integrated into real-time c
 ontrol systems. These different applications of machine learning to magnet
 ic fusion challenges will be addressed in this minisymposium. The talks wi
 ll provide an overview of the advanced ML techniques applied and should th
 us be of interest to a broad audience.\n\nPedestal Inference Engine (PIE):
  ML Facilitated Simulation-Based Inference Framework for Tokamak Pedestals
 \n\nThe increase of readily available computing resources and advancement 
 of simulations-based inference (SBI) algorithms is opening a pathway to br
 ing statistical inference to the forefront in model validation, discovery,
  and prediction in science and technology [1]. In the context of tokamak f
 usion pl...\n\n\nAaro Järvinen, Amanda Bruncrona, Daniel Jordan, Adam Kit,
  and Anna Niemelä (VTT); Laurent Chone (CSC - IT Center for Science); Lore
 nzo Frassinetti and Arnaud Lafay (KTH Royal Institute of Technology); Alex
  Panera-Alvarez, Sven Wiesen, Taweesak Jitsuk, and MJ Pueschel (DIFFER); V
 lado Menkovski and Kiet Bennema ten Brinke (Technical University of Eindho
 ven); Samuli Saarelm and Lorenzo Zanisi (UKAEA); Tobias Görler (IPP Garchi
 ng); Michele Marin (EPFL); and David Hatch (University of Texas at Austin)
 \n---------------------\nSurrogate Models of Nonlinear Gyrokinetics and Hi
 gh-Fidelity Fluid-Kinetic Edge Plasma Codes\n\nHigh-fidelity gyrokinetic t
 urbulence modelling and scrape-off layer modelling are crucial for scenari
 o development, but they not included in integrated models due to the assoc
 iated prohibitive computational cost. Firstly, this talk will present surr
 ogate models of nonlinear gyrokinetics (operating i...\n\n\nLorenzo Zanisi
  (UKAEA); Fabian Paischer (JKU Linz, Mistral AI); Gianluca Galletti (JKU L
 inz, Emmi AI); Gerald Gutenbrunner (JKU Linz); Will Hornsby, Naomi Carey, 
 and Francis Casson (UKAEA); Theodore Brown (UKAEA, DeepMind); Michele Mari
 n (EPFL); Stefano Gabriellini (UKAEA); Enzo Vergnaud and Clarisse Bourdell
 e (CEA); Andrew Snodin, Harry Dudding, and Colin Roach (UKAEA); Garud Snoe
 p and Aaron Ho (MIT); Stanislas Pamela (UKAEA); Johannes Brandstetter (Emm
 i AI); Jonathan Citrin (DeepMind); and George Holt, Ziga Stanca, Romain Fu
 ttersack, and James Simpson (UKAEA)\n---------------------\nScaling and Sc
 arcity: Developing Robust ML Surrogates for Next-Generation Fusion Workflo
 ws\n\nSurrogate modelling is emerging as a key paradigm in fusion research
 . It provides fast, reliable software components to accelerate high-fideli
 ty plasma simulators and digital twins. However, their development is intr
 insically linked to data availability and parameterization. This work expl
 ores data-...\n\n\nSamy Kerboua-Benlarbi (EPFL); Christoph Angerer (NVIDIA
  Inc.); Florian Cabot and Francesco Carpanese (EPFL); Jonathan Citrin (Dee
 pMind); Hammam Elaian (EPFL); Cyrille Favreau (NVIDIA Inc.); Federico Feli
 ci (DeepMind); Davide Fransos (NVIDIA Inc.); Philippe Hamel (DeepMind); Ho
 lger Reimerdes, Cristian Sommariva, and Elena Tonello (EPFL); Felix Yang (
 NVIDIA Inc.); and Alessandro Pau and Olivier Sauter (EPFL)\n--------------
 -------\nThe PORTALS Workflow: Accelerating Core Transport Predictions wit
 h Surrogate-Based Optimization Methods\n\nPredicting core performance in m
 agnetic confinement fusion devices requires accurately modeling the balanc
 e between heating sources and transport mechanisms. However, the nonlinear
  electromagnetic turbulence in tokamak and stellarator plasmas makes such 
 modeling computationally demanding, as capturi...\n\n\nP. Rodriguez-Fernan
 dez and N.T. Howard (MIT Plasma Science and Fusion Center) and J. Candy (G
 eneral Atomics)\n\nDomain: Physics, Computational Methods and Applied Math
 ematics\n\nSession Chair: Stephan Brunner (EPFL)
END:VEVENT
END:VCALENDAR