MIT scientists have completed one of the most demanding calculations in fusion science using a machine learning technique. According to a recently published paper, the method reduced the CPU time required to make the calculations while maintaining the accuracy of the solution. It’s part of a growing effort to use AI to help solve the math and engineering problems of mastering fusion power. “AI is a tool that allows scientists to iterate faster on experiments, make better predictions about how plasma will act in extreme conditions, and build new fusion devices in a more precise manner,” Andrew Holland, CEO of the Fusion Industry Association, told Lifewire in an email interview.
AI Lends a Hand
MIT researchers Pablo Rodriguez-Fernandez and Nathan Howard are working on predicting the performance expected in the SPARC device, a compact, high-magnetic-field fusion experiment currently under construction. While the calculation required a tremendous amount of computer time (over 8 million CPU hours) the researchers managed to reduce the time required. One of the most challenging problems for fusion researchers is predicting plasma temperature and density. In confinement devices like SPARC, the external power and the heat input from the fusion process are lost through turbulence in the plasma. However, MIT researchers used techniques from machine learning to optimize such a calculation. They estimate that the method reduced the number of runs of the code by a factor of four. New research shows that modern AI techniques can be used to control a nuclear fusion reaction, potentially helping accelerate the development of nuclear fusion as a practical power source, Ulises Orozco Rosas, a professor who studies fusion in the School of Engineering at CETYS University in Mexico, told Lifewire via email. He pointed to the AI software that IBM is developing that could be used to control the magnetic fields that contain the plasma inside the tokamak fusion reactor. “The system was able to manipulate the plasma into new configurations that can produce higher energy,” Rosas added.
The Power of the Stars
Fusion promises unlimited, carbon-free energy through the same physical process that powers the sun and the stars. However, the technical challenges of building a practical fusion power plant are formidable and include heating the fuel to temperatures above 100 million degrees and creating plasma. Researchers use strong magnetic fields to isolate and insulate the hot plasma from ordinary matter on Earth. Holland said that building a working fusion power plant will require a detailed scientific understanding of how to confine and initiate a plasma under fusion-relevant conditions—at extreme temperatures or pressures. “While the hardest part is getting plasma into those relevant conditions, the challenges don’t stop there,” Holland added. “The energy will have to be converted into electricity or usable heat; the fuel cycle will have to be built so that the plasma can be sustained over long periods, and the materials of the fusion device will have to be resilient to the extreme conditions within the power plant.” Holland predicted that energy would “revolutionize” the global energy system. Once commercialized and widely deployed, fusion could mean that energy can be produced without pollution, at any time, with no danger to the public or long-lived radioactive waste. It could usher in an era of energy abundance, making energy cheap, always available, and ubiquitous. But Rosas sounded a note of caution, saying that commercial fusion’s success as an energy provider will depend on whether the challenges of building generating plants and operating them safely and reliably can be met in a way that makes the cost of fusion electricity economically competitive. “With increasing concerns over climate change and finite supplies of fossil fuels, better ways must be found to meet our growing energy demand,” Rosas added. “The benefits of fusion power make it an extremely attractive option: no carbon emissions, abundant fuels, energy efficiency, less radioactive waste than fission, safety, and reliable power.”
