University of Washington researchers are leading and collaborating on four research projects as part of Phase 1 in the U.S. Department of Energy (DOE) Genesis Mission, a national initiative to build an AI-for-science ecosystem which accelerates breakthroughs in energy, discovery science and national security.
The DOE awarded a total of 278 Genesis Mission awards for projects that involve more than 300 participating institutions, including DOE and National Nuclear Security Administration national laboratories, universities and companies. As part of the Genesis Mission, awardees will have access to the Genesis Mission Platform, which includes AI frameworks, advanced AI models from industry partners, and high-performance computing resources across DOE’s National Laboratories and partner facilities.
“The Genesis Mission represents the kind of bold, collaborative approach needed to accelerate the complex scientific and technological breakthroughs required in our rapidly changing world,” said UW Vice Provost for Research Mari Ostendorf. “Its emphasis on bringing together diverse expertise and cutting-edge technology with strategic partnerships highlights the important role universities play in driving innovation for the nation.”
The UW-supported Genesis projects span disciplines including advanced sensing technologies, protein design for microelectronic applications, and astronomy data infrastructure, demonstrating the broad potential of AI to accelerate scientific discovery.
Hossein Naghavi, assistant professor of electrical and computer engineering, received a Genesis award to develop neuromorphic terahertz imaging technology for next-generation augmented reality systems.
“Neuromorphic terahertz imagers, or brain-inspired imaging in the terahertz band, give us superhuman vision to see through optically obscured media in real time by combining sensing and computation in a single piece of hardware. The Genesis Mission award supports our research in developing the hardware for the neuromorphic terahertz imager, as well as creating a digital twin of the entire system to train imaging models before implementing them on the actual hardware. This is an important step for future terahertz imaging systems on augmented reality hardware,” Naghavi said.
Building on research conducted in UW’s Terahertz Integrated MicroElectronics (TIME) Lab, the project brings together collaborators at Texas A&M University, the University of Utah, ChipNexus and NVIDIA to integrate sensing and computation directly within imaging hardware. By reducing the amount of data that must be transferred between sensors and processors, the technology could enable lightweight, low-cost augmented reality headsets capable of seeing through visually obscured materials in real time while operating with far lower power consumption than conventional systems.
David Baker, director of the Institute for Protein Design (IPD), and his colleagues are contributing to two Phase I Genesis Mission projects that apply advances in protein design and artificial intelligence to challenges beyond traditional biomedical applications.
One project, BIND (Biophysics-Informed Learning of Coordination for Metalloprotein Design), led by the Lawrence Livermore National Laboratory, seeks to develop an AI framework for designing selective metal-binding proteins. The project leverages quantum chemistry, structural biology, machine learning and high-throughput experimentation to advance the computational design strategies of next-generation rare earth element binders. These advancements will catalyze change in critical mineral recovery, radionuclide management and environmental monitoring, while advancing DOE priorities in predictive biosystems design.
The second project, led by Pacific Northwest National Laboratory, will develop an AI-guided loop for the design, fabrication and evaluation of microelectronic devices. The team will leverage proteins designed to assemble on van der Waals substrates to modulate their properties and organize charge carriers within the synthesized hybrid materials, integrating deep-learning protein models with AI-guided material synthesis and characterization, device fabrication and measurements, and circuit-level performance evaluation to create an iterative design process that improves molecules, materials, and device design.
Both IPD projects demonstrate how UW-developed protein design technologies are expanding into new areas of chemistry, materials science, and advanced manufacturing. “The Genesis Mission award provides necessary support to develop data that will vastly improve our models, an essential next step in delivering advances for key global challenges,” Baker said.
Andy Connolly, director of the eScience Institute and founding director of the DiRAC Institute, and Neven Caplar, a researcher at DiRAC, are partnering with researchers at Carnegie Mellon University to help develop infrastructure that will make it easier for astronomers to combine and analyze data from a wide range of observatories and scientific instruments.
Their project will expand data formats and analysis platforms to support images, spectra, data cubes and other forms of astronomical data while enabling seamless access to information distributed across cloud and high-performance computing systems. The goal is to remove technical barriers that often slow scientific discovery and make large, multimodal datasets more accessible for AI research.
“We want the plumbing to be boring so the astronomy can be spectacular,” Caplar said.