Hexbyte Glen Cove EPFL and DeepMind use AI to control plasmas for nuclear fusion

Hexbyte Glen Cove

Plasma inside the TCV tokamak. Credit: Curdin Wüthrich /SPC/EPFL

EPFL’s Swiss Plasma Center (SPC) has decades of experience in plasma physics and plasma control methods. DeepMind is a scientific discovery company acquired by Google in 2014 that’s committed to “solving intelligence to advance science and humanity.” Together, they have developed a new magnetic control method for plasmas based on deep reinforcement learning, and applied it to a real-world plasma for the first time in the SPC’s tokamak research facility, TCV. Their study has just been published in Nature.

Tokamaks are donut-shaped devices for conducting research on , and the SPC is one of the few research centers in the world that has one in operation. These devices use a powerful magnetic field to confine plasma at extremely high temperatures—hundreds of millions of degrees Celsius, even hotter than the sun’s core—so that nuclear fusion can occur between hydrogen atoms. The energy released from fusion is being studied for use in generating electricity.

What makes the SPC’s tokamak unique is that it allows for a variety of plasma configurations, hence its name: variable-configuration tokamak (TCV). That means scientists can use it to investigate new approaches for confining and controlling plasmas. A plasma’s configuration relates to its shape and position in the device.

Controlling a substance as hot as the sun

Tokamaks form and maintain plasmas through a series of magnetic coils whose settings, especially voltage, must be controlled carefully. Otherwise, the plasma could collide with the vessel walls and deteriorate. To prevent this from happening, researchers at the SPC first test their control systems configurations on a simulator before using them in the TCV tokamak.

“Our simulator is based on more than 20 years of research and is updated continuously,” says Federico Felici, an SPC scientist and co-author of the study. “But even so, lengthy calculations are still needed to determine the right value for each variable in the control system. That’s where our joint research project with DeepMind comes in.”

3D model of the TCV vacuum vessel containing the plasma, surrounded by various magnetic coils to keep the plasma in place and to affect its shape. Credit: DeepMind & SPC/EPFL

DeepMind’s experts developed an AI algorithm that can create and maintain specific plasma configurations and trained it on the SPC’s simulator. This involved first having the algorithm try many different control strategies in simulation and gathering experience. Based on the collected experience, the algorithm generated a control strategy to produce the requested plasma configuration. This involved first having the algorithm run through a number of different settings and analyze the plasma configurations that resulted from each one. Then the algorithm was called on to work the other way—to produce a specific plasma configuration by identifying the right settings.

After being trained, the AI-based system was able to create and maintain a wide range of plasma shapes and advanced configurations, including one where two separate plasmas are maintained simultaneously in the vessel. Finally, the research team tested their new system directly on the tokamak to see how it would perform under real-world conditions.

The SPC’s collaboration with DeepMind dates back to 2018 when Felici first met DeepMind scientists at a hackathon at the company’s London headquarters. There he explained his research group’s tokamak magnetic-control problem. “DeepMind was immediately interested in the prospect of testing their AI technology in a field such as nuclear fusion, and especially on a real-world system like a tokamak,” says Felici.

Martin Riedmiller, control team lead at DeepMind and co-author of the study, adds that “our team’s mission is to research a new generation of AI systems—closed-loop controllers—that can learn in complex dynamic environments completely from scratch. Controlling a fusion in the real world offers fantastic, albeit extremely challenging and complex, opportunities.”

Range of different plasma shapes generated with the reinforcement learning controller Credit: DeepMind & SPC/EPFL

A win-win collaboration

After speaking with Felici, DeepMind offered to work with the SPC to develop an AI-based control system for its . “We agreed to the idea right away, because we saw the huge potential for innovation,” says Ambrogio Fasoli, the director of the SPC and a co-author of the study. “All the DeepMind scientists we worked with were highly enthusiastic and knew a lot about implementing AI in control systems.” For his part, Felici was impressed with the amazing things DeepMind can do in a short time when it focuses its efforts on a given project.

DeepMind also got a lot out of the joint research project, illustrating the benefits to both parties of taking a multidisciplinary approach. Brendan Tracey, a senior research engineer at DeepMind and co-author of the study, says: “The collaboration with the SPC pushes us to improve our reinforcement learning algorithms, and as a result can accelerate research on fusing plasmas.”

This project should pave the way for EPFL to seek out other joint R&D opportunities with outside organizations. “We’re always open to innovative win-win collaborations where we can share ideas and explore new perspectives, thereby speeding the pace of technological development,” says Fasoli.

More information:
Jonas Degrave et al, Magnetic control of tokamak plasmas through deep reinforcement learning, Nature (2022). DOI: 10.1038/s41586-021-04301-9

EPFL and DeepMind use AI to control plasmas for nuclear fusion (2022, February 16)
retrieved 17 February 2022
from https://phys.org/news/2022-02-epfl-deepmind-ai-plasmas-nuclear.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.

% %item_read_more_button%% Hexbyte Glen Cove Educational Blog Repost With Backlinks — #metaverse #vr #ar #wordpress

Hexbyte Glen Cove Consumers might not return to old product choices once finances improve thumbnail

Hexbyte Glen Cove Consumers might not return to old product choices once finances improve

Hexbyte Glen Cove

Credit: CC0 Public Domain

When faced with job losses, a sudden drop in income, or other stormy economic conditions, consumers will likely need to shift their purchasing priorities and preferences. Those changed preferences outlast the contraction and shape choices even after income recovers.

In a series of studies, Penn State Smeal College of Business-led researchers say that consumers may not return to their original spending patterns even after those gloomy economic clouds finally clear up. The findings suggest that, despite their own tight budgets, businesses should continue to reach customers during uncertain economic times, or face negative consequences that stretch beyond the contraction.

The studies build on work seeking to understand how consumers behave in uncertain economic conditions, said Gretchen Ross, a former Penn State Smeal doctoral student in marketing and currently an assistant marketing professor at Texas Christian University.

“There’s this idea that when we have an expanding budget trajectory, we tend to add more categories to the budget, then when we have a decreasing budget trajectory,” said Ross, who was the first author of the paper. “In other words, we spend on more categories on the upward, than the downward. So, we started thinking what happens when you experience a contraction in your budget, but then are able to go back to your original state? For example, what would happen if you lost your job and needed to cut some budget categories, but then you find a new job and go back to previous income levels. Would we go back to spending our income on the same categories?”

According to Ross, consumers typically shift their priorities and preferences during the contraction and those shifts persist after resources are restored. Further, these shifted preferences may be more stable than initially thought.

“As an example, if your income increases, you might start buying fine wine instead of boxed wine,” said Margaret Meloy, professor of marketing and Calvin E. and Pamala T. Zimmerman Fellow, who also serves as the Marketing Department chair. “However, when your budget constricts, going back to boxed wine may feel so aversive that it makes more sense to stop buying wine entirely. In other words, consumers might cut out entire categories of consumption during contractions. When economic resources return, consumers may continue to skip the wine because they have discovered they weren’t enjoying it that much in the first place,” she added.

Judicious marketing

The researchers, who published their findings in a recent issue of the Journal of Consumer Research, one of the top journals in its field, suggest that companies need to watch marketing budgets closely during these contractions.

“You have to prevent your brand from ending up on the cutting room floor when budgets contract,” said Meloy. “If your brand disappears during the contraction, there’s a lower probability that it will return as budgets re-expand. During an economic downturn, it may not be a time to cut your marketing budget; you may want to spend it judiciously on those most likely to cut your brand during the contraction.”

Ross said that companies should also look for ways to help customers manage times of economic struggles.

“There may be ways that companies could help customers during this time,” said Ross. “For example, let’s say I’m experiencing a financial contraction, it’s not that I don’t want to go to Starbucks for a coffee, I just can’t afford it. Perhaps Starbucks could help me by giving me coupons, that might help me stick with the company.”

The researchers conducted several experiments to show that the effect stretched across other domains including time, space and money.

To test a loss and return of resources of time, the researchers recruited 119 people to test how their responses to how they would budget time in a travel scenario. They were asked to allocate time for an original itinerary and then later asked how that itinerary would change if it was shortened and then restored.

Similarly, the researchers recruited 123 participants to explore a loss and restoration of space resources. In this scenario the participants were asked which vegetables they would plant in a garden of 21 rows and then which they would plant it if the space was contracted to seven rows. They were then asked about their plan when the garden was eventually restored to its original dimensions. Did individuals decide to leave some of the vegetables in the initial allocation out of the final allocation across the 21 rows?

To test , the researchers recruited 223 participants to manage a $300 budget that was cut to $100 and then eventually restored back to $300.

“In every domain that you can show a robust effect, it indicates there’s something fundamental to the way you’re forming preferences,” said Meloy.

The team recruited 178 participants for a follow-up study, referred to as a consequential choice study, that tested the preference-forming effect with real resources—in this case, candy.

Finally, the researchers investigated the preference selection of people who faced a real-world example of contraction during the 2018-19 government shutdown.

The researchers said future work may look at how resource contractions affect the saving patterns of consumers once resources are restored. Do people save more after experiencing a ? Another area of research might be to investigate whether preference refinement affects choice satisfaction. For example, researchers could examine how contractions during the COVID-19 pandemic may alter satisfaction with a simpler lifestyle that lasts after the pandemic ends.

More information:
Gretchen R Ross et al. Preference Refinement after a Budget Contraction, Jo

Read More Hexbyte Glen Cove Educational Blog Repost With Backlinks —