Federico Felici and Jonas Buchli, both with backgrounds in fusion research and AI, have launched Fusionality, a startup focused on developing control systems tailored for fusion reactors. The company, based in Lausanne, Switzerland, recently secured $3.7 million in pre-seed funding to build modular hardware and software components that fusion startups can customize for their specific reactor designs.
Fusion power relies on maintaining a superheated plasma where atomic nuclei fuse, releasing energy. Controlling the plasma’s temperature, shape, and fuel levels requires rapid, precise adjustments. While many fusion companies currently build their control systems from scratch, Fusionality aims to leverage the common physics underlying these systems to offer standardized solutions.
Felici and Buchli met while working on AI-driven control for experimental tokamaks, with Felici having experience at EPFL and Google DeepMind, where he developed simulations and machine learning tools for fusion devices. Although AI is expected to enhance fusion control systems, Felici cautions that it is not yet ready to fully manage reactors autonomously.
Initially, Fusionality will focus on magnetic confinement fusion, a leading approach that uses electromagnets to contain the plasma. The startup plans to expand its offerings to other fusion methods over time. The funding will support the development of a select set of modular technologies, envisioned as building blocks that can be combined to meet the diverse needs of fusion reactor operation.
Fusionality’s work addresses a critical gap in the fusion supply chain by providing specialized control solutions, potentially accelerating the path to reliable, grid-connected fusion power.