Data in Context for Responsive Business – OpenGov Asia


The U.S. Department of Energy (DOE) announced $175 million for 68 research and development projects aimed at developing disruptive technologies to bolster the nation’s advanced energy enterprise. Led by the DOE’s Agency for Advanced Research Projects-Energy (ARPA-E), the OPEN 2021 program prioritizes funding for high-impact, high-risk technologies that support new approaches to address energy challenges. clean energy. The DOE’s Argonne National Laboratory received $7.8 million for three projects.

The selected projects — spanning 22 states and coordinated across universities, national laboratories and private companies — will advance technologies in a wide range of areas, including electric vehicles, offshore wind, nuclear storage and recycling. These investments support President Biden’s climate goals to increase domestic clean energy technology production, strengthen the nation’s energy security, and boost the economy by creating well-paying jobs.

Universities, corporations and our national labs are stepping up efforts to advance clean energy technology innovation and manufacturing in America to provide essential energy solutions from renewables to fusion energy to combat the climate crisis. DOE investments show our commitment to empowering innovators to build bold plans to help America achieve net zero emissions by 2050, create well-paying clean energy jobs, and strengthen our energy independence

– Jennifer M. Granholm, U.S. Secretary of Energy

Selected projects will focus on technologies such as the fuel cell revolution for light and heavy vehicles, and technologies to generate less nuclear waste and reduce the cost of fuel. Argonne’s OPEN 2021 project teams include:

Non-neutron transmutation of irradiated nuclear fuel:

This project will develop technology that supports the establishment of commercially viable, dispatchable, and carbon-free nuclear power for the future clean energy market. Partners: Massachusetts Institute of Technology; University of Michigan; University of California, Berkeley; Idaho National Laboratory and Brookhaven National Laboratory. (Amount of scholarship: $3,000,000).

Advanced design of facilities and protective measures enabled by artificial intelligence/machine learning to establish safe and economical recycling of fast reactor fuels:

This project will develop crucial technologies to support the commercialization and licensing of pyroprocessing that enables the recovery and recycling of valuable nuclear materials from advanced reactor spent nuclear fuel. Partner: Oklo. (Amount of scholarship: $3,600,000).

A zero-emission process for the direct reduction of iron by hydrogen plasma in a rotary kiln reactor:

This research aims to disrupt the steel industry by developing a potentially zero-carbon ironmaking process that eliminates the use of coke or natural gas and requires less energy than current processes. Partners: University of Illinois at Urbana-Champaign and ArcelorMittal. (Amount of scholarship: $1,200,000).

Argonne is committed to accelerating solutions to climate change, which will contribute to the prosperity and security of the United States. As part of this effort, our researchers are transforming the way we reuse nuclear fuel, design reactor protection devices and manufacture carbon-free steel. These scientific innovations would not be possible without ARPA-E’s support of clean energy technologies.

As OpenGov Asia reported, in a new collaboration, Argonne computer scientists are leveraging the power of machine learning expertise and supercomputers in the lab. By reducing the number of costly experiments and time-consuming simulations with a new machine learning approach, they can generate accurate models that provide valuable insight into the welding process in significantly less time and at a fraction of the cost.

This approach, called DeepHyper, is a scalable machine learning package developed by Argonne Computer Scientist. Machine learning is a process by which a computer can train itself to find the best answers to a particular question.

DeepHyper automates the design and development of machine learning-based predictive models, which often involve expert-driven trial and error processes. Because no model is the absolute reflection of the truth. Researchers are not primarily trying to find the best predictive model and associated welding condition. Instead, they generate hundreds of highly accurate models, combine them to assess uncertainties in the predictions, and then seek to use those more tested predictions in the manufacturing process.


About Author

Comments are closed.