Skip to main content
Admissions guidance, study banks, and career tools — get started on Shalyam
Research News

AI Framework Controls Fusion Systems Safely in Milliseconds

Mayur Tembhare
  1. Researchers at Princeton University and the US Department of Energy's Princeton Plasma Physics Laboratory developed an AI framework called PACMAN that can make key decisions in fusion systems safely and quickly.

  2. PACMAN was tested on a real fusion system in five experiments and showed promising results.

  3. The fusion reaction can be thwarted by small disturbances in the plasma, known as instabilities, that grow in milliseconds.

  4. Plasma is hard to predict and sophisticated computer programs used to simulate its behavior can take days or months to run.

Topic: Physics

Researchers at Princeton University and the US Department of Energy's Princeton Plasma Physics Laboratory have developed an AI framework called PACMAN that can make key decisions in fusion systems safely and quickly. The framework was tested on a real fusion system in five experiments and showed promising results.

Inside some fusion energy systems, particles hotter than the core of the sun can become unruly in a few thousandths of a second. This is too fast for human operators to react.

A new software framework called PACMAN has been developed by researchers at Princeton University and the US Department of Energy's Princeton Plasma Physics Laboratory. PACMAN uses artificial intelligence (AI) to make key decisions in fusion systems safely and quickly. The AI framework was successfully tested on a real fusion system in five experiments. The fusion reaction can be thwarted by small disturbances in the plasma, known as instabilities, that grow in milliseconds. Plasma is also hard to predict. The sophisticated computer programs used to simulate the behavior of a plasma can take days or months to run.

The PACMAN framework works like an assembly line with four stations. First, it gathers all the tokamak's real-time measurements, including temperatures, densities and magnetic signals. Then it checks those values against what is expected and makes adjustments as needed. The result is a control loop involving multiple machine learning models that operates at a pace no human could match.

PACMAN was developed by graduate students Hiro Farre Kaga and Andy Rothstein, who are part of the Princeton Program in Plasma Physics. They wanted to create a framework that would allow models to communicate with each other and share outputs. This would enable researchers to do exciting physics in one integrated system.

Why It Matters

Fusion energy could be a virtually unlimited source of electricity. If we can perfect the process, it could provide clean and sustainable energy for India's growing population.

Key Facts

  • Researchers at Princeton University and the US Department of Energy's Princeton Plasma Physics Laboratory developed an AI framework called PACMAN that can make key decisions in fusion systems safely and quickly.
  • PACMAN was tested on a real fusion system in five experiments and showed promising results.
  • The fusion reaction can be thwarted by small disturbances in the plasma, known as instabilities, that grow in milliseconds.
  • Plasma is hard to predict and sophisticated computer programs used to simulate its behavior can take days or months to run.
  • PACMAN works like an assembly line with four stations and uses multiple machine learning models to make decisions.

Key Terms

Artificial Intelligence (AI)
A type of computer program that can learn from data and make decisions on its own.
Machine Learning
A type of AI that uses algorithms to analyze data and make predictions or decisions.

Implications

Fusion energy could be a virtually unlimited source of electricity. If we can perfect the process, it could provide clean and sustainable energy for India's growing population.

Source: https://phys.org/news/2026-09-pacman-ai-framework-fusion-safely.html

Leave a comment

← All research

Get Exclusive Insights with Every Issue

Join Shalyam Newsletter

Stay ahead in education, research, and innovation—straight to your inbox.