July 2017

TAE Technologies and Google Combine Human and Machine Interaction to Further Plasma Science

“Optometrist Algorithm,” First Described Today in the Journal Scientific Reports, Identified Optimal Fusion Conditions

In a paper published today in the journal Scientific Reports, researchers from TAE Technologies, the world’s largest private fusion company, and Google have jointly furthered the understanding of physics and computational design, applying a new approach to solving highly complex problems regarding plasma behavior. The findings will help TAE achieve optimal conditions for fusion energy, and may offer future benefits to other complex, variable areas of research.

The achievement is the first publicly-announced result of a multi-year relationship between TAE and Google, which began in 2014 to apply machine learning to advance plasma research.

Plasma science is a challenging field of study as its subject exhibits complicated behaviors and relies on engineering mechanisms that themselves pose variable conditions and input parameters. Experimental by nature, plasma fusion devices require human judgement to assess results and ensure the operations of one-of-a-kind machinery, further limiting traditional test capabilities.

To increase the speed of learning within these boundaries, researchers from the two companies innovated upon a stochastic perturbation method, one similar to those used in large-scale population and atmospheric models, by incorporating an element of human choice. The team’s “Optometrist Algorithm,” like an eye exam, requires researchers to choose between successive pairs of possible outcomes identified within the experiment to focus on those producing subjectively better results for fusion experiments.

This groundbreaking technique led to the discovery of unexpected plasma confinement results increasing net heating power within TAE’s proprietary field-reversed configuration plasma generator, a more than 50% reduction in energy loss rate and concomitant increase in ion temperature and total plasma energy. Experimental settings specifically revealed particle beam heating led to sustained ion temperature increases, marking significant improvements in plasma containment and stability.

“Results like this might take years to solve without the power of advanced computation to rapidly scale our understanding of the complex properties of plasma. This research, years in the making, has already allowed us to advance our science, and I am grateful for the longstanding collaboration with Google as it illustrates how, with a lean start-up design, TAE can direct research to outcomes that specifically advance and accelerate our mission to generate clean, safe and abundant energy through fusion technology,” said Michl Binderbauer, TAE’s President and Chief Technology Officer.

Read the full paper.

Building on the success of the Optometrist Algorithm, Google and TAE will continue to collaborate on Norman, the company’s new plasma generator, which initiated operations last month.

About TAE Technologies
TAE Technologies (pronounced T-A-E) was founded in 1998 to develop commercial fusion power with the cleanest environmental profile, and represents the fastest, most practical, and economically competitive solution to bring abundant energy to the grid. With over 900 issued patents, more than $880 million in private capital, six generations of National Laboratory-scale devices, and an experienced team of over 250 employees, TAE is now on the cusp of delivering this transformational energy source capable of sustaining the planet for centuries. The company’s revolutionary technologies have produced a robust portfolio of commercial innovations in other large adjacent markets such as power management, energy storage, transmission, electric mobility, life sciences, and more. TAE is based in California, and maintains international offices in the UK and Switzerland. Multidisciplinary and mission-driven by nature, TAE is leveraging proprietary science and engineering to create a bright future for us.

Sign up for our newsletter.