Two former Google DeepMind and EPFL researchers, Federico Felici and Jonas Buchli, have identified a critical gap in the burgeoning fusion energy sector: the lack of standardized control systems for experimental fusion devices. Recognizing that many fusion companies are forced to develop these complex systems from scratch, the duo founded Fusionality this year, aiming to provide specialized hardware and software solutions.

Felici and Buchli’s extensive experience in developing control mechanisms for experimental fusion reactors, including work on tokamaks and machine learning interfaces for fusion devices, positioned them uniquely to address this industry-wide challenge. They observed that while 80% of a control system’s core functionality is consistent across different reactor designs, companies individually invest significant resources in redundant development. Fusionality seeks to streamline this process by offering a suite of adaptable control systems and simulation environments that can be fine-tuned for specific reactor designs.

The Lausanne-based startup recently secured a $3.7 million (CHF 3 million) pre-seed funding round from Founderful and Playfair. Felici, serving as CEO, and Buchli, as CTO, lead a team of seven, focusing initially on magnetic confinement fusion, an approach utilized by companies like Commonwealth Fusion Systems, Realta Fusion, Proxima Fusion, and Type One Energy. While AI is expected to play an increasingly important role in optimizing parts of future control systems, Felici emphasizes that it will primarily complement, rather than fully control, the intricate operations of a fusion reactor.

Fusionality’s strategy involves developing a select number of foundational technologies, described as ‘Lego blocks,’ which can be expanded to serve the diverse technological needs of operating a fusion reactor. By standardizing these critical components, the company aims to accelerate the development and deployment of fusion power plants, moving closer to a future where fusion energy contributes to the grid.

Key Takeaways

• Fusionality, founded by Google DeepMind and EPFL alumni, aims to standardize control systems for fusion reactors, addressing a critical industry gap.
• The company secured $3.7 million in pre-seed funding to develop adaptable hardware and software solutions, initially focusing on magnetic confinement fusion.
• Fusionality's approach seeks to reduce redundant development efforts by providing common control system components, accelerating the path to grid-scale fusion energy.

Editor’s Analysis & Impact

Fusionality’s entry into the fusion energy sector marks a significant step towards industrializing a technology often seen as decades away. By focusing on standardized control systems, the company addresses a crucial bottleneck: the high cost and complexity of custom-building essential operational components for each new reactor design. This standardization could dramatically reduce development timelines and capital expenditure for fusion startups, fostering a more efficient and collaborative ecosystem. The involvement of AI, even in a complementary role, highlights the industry’s embrace of advanced technologies to manage the extreme conditions within fusion reactors. If successful, Fusionality could accelerate the commercial viability of fusion power, impacting global energy markets and the transition to clean energy sources.

Frequently Asked Questions

Q: What problem is Fusionality trying to solve in the fusion energy industry?
A: Fusionality aims to solve the problem of fusion companies having to build their control systems from scratch. They are developing standardized hardware and software solutions that can be adapted for various fusion reactor designs, thereby reducing development costs and time.

Q: Who founded Fusionality and what is their background?
A: Fusionality was founded by Federico Felici (CEO) and Jonas Buchli (CTO). Both have extensive experience in developing control systems for experimental fusion devices, with backgrounds at institutions like Google DeepMind and EPFL.

Q: How will AI be integrated into Fusionality's control systems?
A: While AI is not expected to fully control fusion reactors, Fusionality anticipates it will play an important role in future control systems by complementing, enhancing, or optimizing specific parts of the system, such as maintaining plasma conditions.