From Multiphysics Simulation to Immersive Scientific Analysis: A Procedural Data Processing Framework for Autonomous Virtual Laboratories with Active Thermography as a Case Study

This paper proposes a simulation-driven data processing framework that transforms multiphysics simulation outputs into interactive representations suitable for immersive scientific visualization and analysis. The proposed workflow uses Comsol Multiphysics 6.1 software for numerical simulation, Houdini FX for procedural preprocessing and spatial reconstruction, and Unreal Engine for real-time visualization and interaction with standalone Meta Quest headsets. Houdini imports point-based numerical datasets, reconstructs their spatial structure, and performs interpolation and filtering operations. It then exports an optimized representation that is compatible with real-time visualization. Unreal Engine then provides dynamic visualization, temporal navigation, temperature-scale adjustment, graphical analysis tools, and spatial selection of regions of interest. Active thermography is employed as a case study because it generates spatially distributed, time-dependent thermal fields that require the preservation of spatial and temporal information throughout the processing workflow. The results demonstrate the technical feasibility of connecting multiphysics simulations, procedural data processing, and standalone, immersive visualization within a modular workflow. Although evaluated here through active thermography, the architecture is extensible and provides a basis for future investigations with other point-based scientific simulation and experimental datasets.