Advanced techniques for fusion data visualisation

Abstract

The research and development of fusion energy technology is becoming increasingly complex and involving large amounts of diverse data like engineering CAD models, multi-physics simulations, and AI-based diagnostics. Such diversity and multi-modality of data necessitate the creation of advanced visualisation systems for improved data integration, accessibility, and usability. This study explores the adaptation of advanced visualisation techniques to fusion data through three core domains: (1) efficient analysis and instinctive exploration of heterogeneous datasets; (2) construction of visualisation pipelines enabling iterative refinement and retrospective analysis; and (3) deployment of new tools and technologies tailored to fusion-specific applications such as 3D visualisation, real-time dashboards, and immersive environments. As a result, this paper presents an integrative approach to combining diverse fusion data sources using advanced tools such as NVIDIA Omniverse, ParaView, Blender, Grafana, and WebXR. We further discuss a framework integrating simulation data, diagnostics, and design models into an interactive ecosystem. We demonstrate its effectiveness through key use cases, including camera-like MHD simulations, interactive diagnostic dashboards, and immersive AR/VR visualisation of tokamak data. These advances enhance scientific understanding, facilitate cross-disciplinary collaboration, and pave the way for future AI-driven adaptive visualisation in fusion research.