Michael Keidar, the A. James Clark Professor of Engineering at the George Washington University School of Engineering and Applied Science, has secured an inaugural award from the U.S. Department of Energy’s (DOE) Genesis Mission. The national initiative funds projects that deploy artificial intelligence (AI) into research and development workflows to accelerate discovery and spark innovation. Keidar’s was among the less than 6% of submitted proposals that received funding.
“It’s very exciting,” Keidar said. “This is a flagship program to involve artificial intelligence in all different levels of the way that we do science in the U.S.”
“Professor Keidar's Genesis Mission award is a testament to GW’s growing expertise in AI and continued commitment to research that has real-world impact,” Provost Edward Balleisen said. “As we implement the Strategic Framework, we will be expanding our capacity to develop this kind of ambitious interdisciplinary team, which so often requires long-term partnership with companies and other external organizations.”
Keidar’s work centers on low-temperature plasmas (LTPs), particularly their uses in satellite propulsion and biomedicine. Plasma is a form of matter in which a gas’s neutral atoms have been separated from their negatively charged electrons, resulting in free electrons and positively charged particles (ions). This composition makes plasmas electrically conductive, behaviorally complex and—in theory—endlessly flexible through the application of electric and magnetic fields.
LTPs can be applied to human tissue for an expanding range of therapeutic purposes. One type of plasma could deliver nitric oxide, a molecule known to accelerate wound healing; another might selectively kill cancer cells without damaging healthy tissue. In every case, AI-driven therapeutic devices will monitor and analyze a patient’s individual response to this “adaptive plasma” and instantly adjust their treatment.
“The uniqueness of plasma is that you can actually modify it in real time,” Keidar said. “You can change the settings of the plasma device and almost instantaneously change the entire chemistry.” The theoretical result, he explained, is a “highly personalized” therapy, especially compared with pharmaceutical drugs, whose chemical makeup is “set in stone.”
This advantage of plasma doesn’t just apply to biomedicine. Theoretically, the same plasma production and control system could also provide responsive, tailored solutions in agriculture, manufacturing, food processing, environmental remediation and beyond.
AI can help close the gap between theory and practice, Keidar said. Machine-learning models emulate the complex physics of plasma processes, helping Keidar and his interdisciplinary team develop scalable, high-precision LTP platforms—perhaps even, eventually, autonomous platforms that could respond automatically to feedback from a patient or system without the need for an expert monitor.
In the first stage of their Genesis project, Keidar and his team will use machine learning to evaluate the most effective plasma chemistry for three different treatments—cancer therapy, wound healing and decontamination—and explore the potential for a single smart device to be used for all three. In the process, they’ll create updated algorithms that could enable even more powerful approaches. Their results will be externally validated by the DOE’s Princeton Plasma Physics Laboratory. Industry partner Micro1 will help in developing advanced machine learning algorithms.
Keidar’s interdisciplinary project team includes coinvestigators Li Lin, GW research scientist; Sophia Gershman, Yevgeny Raitses and Shurik Yatom of the Princeton Plasma Physics Laboratory; and Hunter Hayden of Micro1.
Ultimately, Keidar and his team hope to develop optimized, autonomous adaptive plasma platforms, powered by AI and adaptable across a range of fields and industries.
Already, Keidar said, he’s seeing exciting possibilities develop. Major companies like Google and Amazon, among many others, participated in last week’s Genesis Mission Summit and have expressed interest in collaborating with researchers.
“People are very excited to be a part of this,” Keidar said.