The missing link in fusion

DeepMind alumni Federico Felici and Jonas Buchli launched Fusionality after realizing fusion startups were all struggling to build reactor control systems from scratch.

Fresh capital for control systems

The Lausanne-based startup raised $3.7 million in pre-seed funding from Founderful and Playfair to build standardized tools for fusion reactors.

Ending duplicated engineering effort

While an emerging supply chain provides specialized hardware components, around 80% of every fusion company's control system relies on the exact same underlying logic.

“Across the industry, many companies make their own control systems from scratch. But actually, 80% of each company’s control system is really exactly the same.”

Taming fickle plasma in real time

Fusion power relies on maintaining superheated plasma at precise temperatures and shapes. Fusionality is creating core control systems and simulation software startups can fine-tune.

Rooted in DeepMind research

The founders previously teamed up to teach AI how to control a tokamak, a donut-shaped fusion device, at EPFL and Google DeepMind.

AI as assistant, not driver

While artificial intelligence will play a critical role in optimizing reactor performance, the team believes AI should complement core physics controls rather than run whole reactors alone.

“I wouldn’t be somebody who advocates for AI to take the role of controlling the entire fusion reactor.”

Targeting magnetic confinement first

Fusionality is launching its software for magnetic confinement devices, the approach used by startups like Commonwealth Fusion Systems, before expanding to other reactor types.

Building modular Lego blocks

The seven-person team is focusing on a tightly selected set of modular technology components with the goal of eventually powering a wide range of fusion operations.

“We start with a few of these blocks and then we build out more.”