The grimoire of the fusion machine

Digital twins are becoming central to fusion engineering as teams work to integrate simulation, manufacturing data and operational insight across the full lifecycle of high value components. Recent work has shown how digital twins can create a complete digital thread in fusion manufacture, linking CAD models, robotic build instructions, in process sensor data, geometric scans and physics based simulations into a single, continuously updated environment. At the same time, the fusion community is converging on the idea of digital twins as living platforms that continually ingest new experimental data, refine predictive models and support more informed engineering decisions. This aligns closely with the FAIR data principles, which promote the Findability, Accessibility, Interoperability and Reusability of data across organisational boundaries.<br/><br/>In this work we present early progress toward a FAIR aligned digital twin infrastructure designed specifically for fusion engineering. The system, currently under active development, aims to capture multi modal data with complete provenance and version control, enabling rigorous traceability and consistent reuse of information across design, manufacture and operation. To illustrate the vision behind this effort, we introduce the metaphor of “The grimoire of the fusion machine.” In folklore, a grimoire is a living book of knowledge that grows with every entry. Here, the term reflects the goal of creating an evolving, authoritative archive of engineering evidence. Unlike its mythical namesake, this “grimoire” would be grounded in engineering rigour: entries generated automatically as data passes through measurement pipelines, reference measurements from physical testing, and simulation workflows, forming a transparent and auditable lifecycle record.<br/><br/>Although the infrastructure is still a work in progress, its emerging framework demonstrates how FAIR oriented design can support verification ready simulation, improved uncertainty management and clearer decision making for safety critical fusion components. For industrial practitioners, the approach points toward future digital engineering environments where information flows more freely, integration friction is reduced and AI assisted analysis becomes increasingly viable.