Back to Bigger Picture

BIGGER PICTURE Energy & climate

How an AI learned to hold two plasmas in one vessel

You cannot hold fusion plasma with an ordinary clamp. At EPFL, researchers trained a controller to shape it with magnetic fields—and then trusted it with a real experiment.

Purple plasma glows inside the TCV tokamak vessel.Original photograph
Plasma inside TCV. The research challenge is to control its position and shape using magnetic fields.2022 EPFL / Curdin Wüthrich – SPCCC BY-SA 4.0
Real experimental machine
TCV · Lausanne
Demonstrated configuration
Two simultaneous plasmas
What was demonstrated
Magnetic control, 2022

How the idea works

  1. Specify a plasma shape

    Researchers define the configuration the experiment needs.

  2. Practise coil adjustments

    A simulated tokamak rewards settings that approach the target.

  3. Control the real coils

    The trained policy shapes plasma in TCV.

The invisible grip around the plasma

A tokamak confines extremely hot, electrically charged gas with magnetic fields. Coils around the vessel influence its position and shape. Letting it drift into the walls loses heat and can damage the machine. TCV, the variable-configuration tokamak in Lausanne, is particularly useful because researchers can explore different shapes rather than operating with just one fixed geometry.

Sources: [1]

Practise first in a simulated reactor

The controller learned through reinforcement learning: try coil settings in a physics simulation, evaluate how closely the resulting plasma matches a requested configuration, then improve the strategy. DeepMind describes the real control problem as adjusting voltages thousands of times per second. Training can be lengthy; the learned controller must make each operating decision quickly enough for the physical system.

Sources: [3]

Cutaway scientific illustration of the TCV vessel, plasma and surrounding magnetic coils.Source illustration
A model of the vessel and its coils: the controller changes the coil voltages to shape the plasma. Scientific illustration.2022 EPFL / DeepMind and SPCCC BY-SA 4.0

The surprising test: two separate plasmas

In the 2022 experiment, the trained system produced and maintained several configurations on the real TCV machine. One held two separate plasmas simultaneously inside the vessel. This is useful experimental flexibility: physicists can investigate how different configurations behave. The achievement was controlling the apparatus and its plasma, not generating electricity for a household or demonstrating a commercial fusion power station.

Sources: [1][2]

The useful bridge from practice to hardware

The broader lesson lies in the handover from simulated practice to real hardware. Researchers supplied physical models, target configurations and an experimental facility; learning supplied a control strategy. Better experimental tools may accelerate energy research, but they do not remove the remaining physics and engineering problems. The human benefit here is a research capability, with a long route to any electricity supply.

Sources: [1][3]

The original sources

  1. EPFLEPFL and DeepMind use AI to control plasmas for nuclear fusion
  2. NatureMagnetic control of tokamak plasmas through deep reinforcement learning
  3. Google DeepMindAccelerating fusion science through learned plasma control

What to explore next

Why NASA’s AI-designed parts look like bonesMillions of imagined crystals, one real battery
Explore all 99 applications