Blueprints - Webflow Ecommerce website template

What is Nuclear Fusion?

‍

Nuclear Fusion could be defined as “the process by which two light atomic nuclei combine to form a single heavier one while releasing massive amounts of energy” (Mateo). In simpler terms, fusion is the process of combining two or more atoms into a single, heavier atom with energy being expelled at the same time.

‍

So, why is this important? Nuclear fusion is an important concept because it could be the solution to creating clean, limitless energy to power everything on Earth. In comparison to nuclear fission or the splitting of an atom to create energy, nuclear fusion could generate “four times more energy per kilogram of fuel than fission” or in other words “nearly four million times more energy than burning oil or coal” (Mateo). Not only that, but unlike fission reactions which require a hard-to-find mineral, uranium, and output of radioactive waste, fusion could be done using the isotopes tritium and deuterium from hydrogen in seawater with an output of a finite resource, helium.

‍

Overall, nuclear fusion is vastly superior to fission due to its potential for limitless, clean energy.

‍

The Tokamak Nuclear Reactor

‍

The Tokamak Nuclear Reactor is one of the leading technological advances made in nuclear fusion, which takes the approach of heating a ring of gas known as a plasma inside a torus-shaped tube.

‍

This donut-shaped device is designed to confine the reactants while maintaining an incredibly strong magnetic field. So, why aren’t we using fusion nuclear reactors now to create energy?

‍

There are many reasons why nuclear fusion reactors such as the Tokamak are not currently being used to supply the world with energy. For one, “magnetic fields struggle to contain plasmas that reach above 100 million degrees Celsius, hotter than the center of the Sun” (Poore and Center).

‍

In essence, simply heating the jelly-like plasma is a momentous task. Furthermore, while keeping the core of the plasma hot enough for fusion to occur is difficult, this must also be achieved while keeping the walls from melting from the heat. However, that’s not to say that it isn’t possible.

‍

Amazing milestones have been achieved that have shocked the world. For instance, in September of 2022, “The Korea Superconducting Tokamak Advanced Research sustained plasma at ion temperatures hotter than 100 million kelvin for 30 seconds” (Seo et al.). Not only that but in December of 2021, “the Joint European Torus broke the world record by producing 59 megajoules of fusion energy for 5 seconds (Seo et al.). These amazing achievements are reminders that steps are being taken in the direction of clean, limitless energy. However, while the Tokamak has made undeniable progress in the pursuit of fusion energy, there is still another, even more difficult problem to solve regarding plasma.

‍

The Real Issue of Plasma

‍

We’ve already covered how it’s difficult to achieve the optimal heat for the plasma and how it's difficult to keep the walls from melting down during a nuclear reaction. However, there is another major issue with plasma, which is the focus of the Princeton team's efforts.

‍

This problem is known as a plasma disruption event. This occurs when plasma tears and “[escapes] the machine’s powerful magnetic fields that are designed to keep the plasma contained” (Brennan).

‍

This is an extremely difficult problem to prepare for as predicting instability before they happen requires inhuman reactions. This is where Princeton’s use of AI, comes into play as the next milestone toward fusion energy.

‍

Princeton’s Use of AI

‍

“Since tearing mode instabilities can form and derail a fusion reaction in milliseconds, the researchers turned to artificial intelligence for its ability to quickly process and act in response to new data'' (Poore and Center).

‍

Princeton’s research team turned to deep reinforcement learning (RL) as the technique for performing nonlinear, high-dimensional actuation problems to solve the issue of the tearing plasma. Utilized in the Tokamak, the “RL algorithm [would] optimize the actor model based on deep neural network (DNN), and the actor model gradually learns the action policy leading to higher rewards in the given environment” (Seo et al.).

‍

In summation, Princeton’s team used a type of A.I model that would learn through success to actively control the Tokamak to keep a high-pressure plasma whilst keeping the possibility of tearing low. They achieved this goal by training the RL in an environment that employs a dynamic model that predicts future plasma pressure and tearing likelihood to develop an A.I controller that would actively control the actuators to keep high-pressure plasma while maintaining the low tear ability in the plasma (Seo et al.).

‍

What’s the big deal about this AI model? The reason why Princeton’s AI is such a critical milestone is that it can predict plasma instabilities before they happen by just three milliseconds. This small amount of time is just enough for an AI controller to take corrective action with the Tokamak’s actuators before the plasma is disrupted. Essentially, Princeton’s AI is one of the first steps to keeping plasma from tearing long-term meaning that a fusion nuclear reactor has the potential to be everlasting with the use of deep learning models in the future.

‍