Less waiting.
More exploration.
We aim to unite established fusion simulation codes with accelerated numerical methods, shortening the time between a design question and a physics-grounded answer.
Connect the physics.
Specialised codes address different parts of fusion physics. Our aim is to connect these methods in a self-consistent modelling workflow, so changes in geometry, heating and plasma behaviour can be evaluated together.
Faster calculations and tighter coupling would let us explore more physically consistent combinations, compare design alternatives and understand their effects across the reactor.
Speed with physical fidelity.
We aim to accelerate numerical computation within established, high-fidelity physics methods. Accuracy, convergence and physical consistency set the conditions for usable speed.
AI-assisted exploration and physics-aware surrogate models may support parts of the workflow, with their outputs checked against the relevant numerical and physical evidence.
Modelling serves the plant.
The goal is to discover and evaluate reactor geometries, guide physical engineering and ultimately build a fusion power plant. Integrated modelling is the means of reaching that goal.
Digital twins are a possible supporting use of this modelling: a way to study plant behaviour and test operational ideas before full-power operation. They could also inform future monitoring and maintenance. These remain research aims.
Reactor research