Summary of the project

The project will build a diverse set of virtual laboratories (V-Labs) and a digital platform that will enable the design and performance testing of fusion energy systems—systems that use fusion reactions for power generation and various other applications—in a digital space.
To reproduce, in the digital domain, the state of plasma (ionized gas) in fusion energy systems, as well as the complexity and spatiotemporal scales (temporal and spatial extents) of their components, we will define a new “Super Dimensional Data Space” that integrates axes such as time, space (coordinates), velocity, and physical quantities. Based on this uniquely structured space, we will establish computational methods that incorporate its characteristics.
Furthermore, to apply these computational methods as “Super Dimensional State Engineering,” we will develop innovative AI and data‑driven scientific technologies. By integrating such advanced AI/data-driven science with knowledge from mathematical sciences, fusion energy science, and other disciplines, we will construct V-Labs that enable performance predictions—future forecasting—of component technologies and entire next‑generation fusion energy systems within a digital space.
To further promote the societal deployment of the V-Labs and to dramatically enhance the accuracy of their performance predictions, we will establish a new “Fusion Data Space,” a massive virtual library in the digital space. This platform will enable sharing of simulation programs used in fusion energy science, experimental data, computational data, documentation, and more among trusted researchers

Fig. 1. Conceptual Illustration of Next-Generation Fusion Energy System Design Using the V-Lab.

Through this project, we aim to significantly reduce the time‑ and cost‑intensive trial‑and‑error processes—such as prototype development and performance testing—that are required in the real world. By doing so, we will accelerate the early social implementation and cost reduction of various fusion energy systems, ultimately contributing to the realization of a society powered by fusion energy.

Milestone by year 2034

We will further expand the database of the “Fusion Data Space,” which is linked to the versatile V-Lab (the Generic V-Lab) that incorporates advanced AI/data driven science and mathematical science, elevating both the volume of data and the diversity of information to a world leading level.
In addition, using the Generic V-Lab as a common foundation, we will conduct benchmarking in three types of V-Labs constructed in the digital space—(1) magnetic confinement fusion systems, (2) non magnetic confinement fusion energy systems, and (3) material testing facilities such as neutron irradiation environments—by using experimental results from existing real world devices. Through these efforts, we will demonstrate that all three V-Labs can reproduce the experimental behavior of existing real world devices.

Milestone by year 2029

We will construct the Generic V-Lab and integrate it with the newly established Fusion Data Space. Using the Generic V-Lab as a common platform, we will perform benchmarking in the three types of V-Labs constructed in the digital space and demonstrate that one of the three V-Labs can reproduce the experimental results obtained from existing real world devices.

R&D theme progress reports

• [1] Generic V-Lab
• [2, 3, 4] V-Lab A, B, C
• Progress Report (366 KB)

Performers

PDF Download

• Summary of the project (263KB)