Progress Report

Backcasting Digital Systems by Super Dimensional State Engineering[1] Generic V-Lab

12, 3, 4»

Progress until FY2025

1. Outline of the project

In this R&D theme, we are constructing a versatile V-Lab (the Generic V-Lab), which will serve as the foundation for virtual laboratories (V-Labs)—digital-space experimental environments—that can be applied across a wide range of fields. To achieve this, researchers from diverse disciplines, including AI and data‑driven science, computational physics, mathematical sciences, and high‑performance computing, are collaborating to develop methods for integrating experimental data, theories and models, and computational (simulation) data (Fig. 1). The Generic V-Lab is being realized by extending the “Open Data Analysis Tool for Science and Engineering (ODAT SE),” a general-purpose data analysis framework developed previously by the PM (Hoshi) et al.

Fig.1
Fig. 1. Conceptual illustration of the integration of experiments, theories and models, and computational data that forms the foundation of the Generic V-Lab.

2. Outcome so far

In FY2025, we primarily conducted research on essential technologies for the Generic V-Lab, including AI and data-driven scientific approaches and high-speed mathematical algorithms. We also developed a prototype of the Generic V-Lab by extending ODAT SE and demonstrated its usefulness through applied problems (examples are shown in Fig. 2).

Fig.2
Fig. 2. Demonstration of the usefulness of the Generic V-Lab applied to Thomson scattering diagnostics: schematic image of the experiment (top), and comparison between measured (synthetic) scattering spectrum and the estimated spectrum (bottom).

The Generic V-Lab must be capable of supporting a wide variety of experiments simply by incorporating programs tailored to different experimental devices. In Fig. 2, we demonstrate its usefulness using the “Thomson scattering diagnostic system,” which is widely used in fusion energy research. This diagnostic system measures the velocity distribution of particles within a plasma. Our results show that the system can accurately determine particle velocity distributions even under complex conditions where particle motion is not uniform. Although ODAT‑SE was originally developed for materials science, it has been expanded through this project to form the foundation of the Generic V-Lab, enabling its application to a wide range of experimental devices across multiple fields. This is the first time the framework has been applied to instruments in plasma science, and the work has demonstrated the versatility required for constructing various V-Labs.

3. Future plans

We will further advance the data integration methods for experimental data, theories and models, and computational (simulation) data (Fig. 1), thereby enhancing the capabilities of the Generic V-Lab. In parallel, we will collaborate with the principal investigators (PIs) of Research and Development Themes 2 through 4 to develop V-Labs tailored to diverse fusion energy systems as well as material experiments such as neutron irradiation (details on V-Lab construction are provided on the next page). As the project progresses, we plan to sequentially add new research and development topics and PIs.

12, 3, 4»