Nuclear engineer uses artificial intelligence to model nuclear fallout dispersion Meet Nathan Boyle

Nuclear engineer uses artificial intelligence to model nuclear fallout dispersion

Using machine learning, Nathan Boyle hopes to create more accurate tracking models for radioactive-chemical plumes and make the world safer for nuclear energy. (Photo Credit: Nathan Boyle)

Nathan Boyle found his passion for science, technology, engineering and mathematics (STEM) while growing up in a small town in Indiana, where he tutored his classmates and family friends. During high school, his older brother gifted him a hand-me-down copy of “Introduction to Nuclear Engineering.” He has been fascinated by the subject ever since.

Boyle earned his bachelor’s and doctoral degrees in nuclear engineering from Purdue University. As a postdoctoral student, he served as a mentor to more than 30 high school, undergraduate and graduate students which gave him the opportunity to share his love of teaching and tutoring.

After graduation, Boyle made a proposal for the Office of the Director of National Intelligence’s (ODNI) Intelligence Community Postdoctoral Research Fellowship program (IC Postdoc). He was accepted and became a fellow studying under his mentors Rusi Taleyarkhan and David Stout as a part of the Metastable Fluids and Advanced Research Laboratory at Purdue University. The IC Postdoc Program offers scientists and engineers from a wide variety of disciplines unique opportunities to conduct research relevant to the intelligence community.

His research project focuses on nuclear fallout. Specifically, Boyle is studying artificial intelligence (AI) machine learning models designed to predict and simulate the spread of nuclear fallout, called plume modeling. As the world shifts toward nuclear energy and for the possible deployment of nuclear weapons, it becomes imperative to have accurate intelligence-related tracking and emergency preparedness measures.

Boyle hopes to make these models more accurate by using open-source images, real time drone verification and optimized AI inputs. One example includes using a piece of code created by the Sandia National Laboratories called Melting Core (MELCOR) to model the progression of nuclear accidents with the Accident Consequence Code System (MACCS).

Using data from a real event demonstrates the effectiveness of his model. “Results of simulating the 1986 Chernobyl nuclear catastrophe in Ukraine compared well with actual published measurements,” explained Boyle.

“The idea is to preemptively predict the dispersion of release events that may occur. We do that by developing algorithms that have a robust and verified output for simulated dispersions. This will allow for quicker decisions, which will allow for quicker evacuations to save lives,” he continued.

Boyle’s goal is to contribute to the advancement of nuclear safety and security and to meet the environmental needs of the country. These simulations and machine learning models can be used by intelligence, first responder and safety-related industries to save lives and achieve those goals. By making more accurate predictions, leaders can mitigate the damage or potential damage caused by nuclear emergencies.

Boyle presented the research at the 8th Annual Intelligence Community Academic Research Symposium. He received the School of Nuclear Engineering Distinguished Dissertation Award, Outstanding Student Research Award and the Bilsland Dissertation Fellowship Award.

As a fellow, Boyle’s skill set is growing as he learns more about AI and simulation tools. His daily schedule usually includes spending time with the machine. He creates and simulates data and then optimizes the machine to be more accurate. He believes his new experiences with AI will be beneficial as the technology becomes a larger part of general society. His favorite part of the fellowship is being allowed to develop his own research goals.

Boyle recommends the IC Postdoc program.

“This program has allowed me to integrate my knowledge of simulation stools in nuclear engineering and adapt them toward plume monitoring for release events,” said Boyle. “This program also allows me to have the academic freedom to adapt my research focus on what I believe is the best approach and strategy toward completing my research goals.”

After his fellowship, Boyle plans to transition into the industry and become a professor, where he can combine his love of teaching with his love of STEM. Additionally, Boyle wants to continue collaborating on research projects and address gaps in technology such as integrating the machine learning plume models where they would be most effective and other gaps such as novel sensors for nuclear medicine.

For now, Boyle will remain an IC postdoc fellow and researcher with the ODNI, where he will continue improving the accuracy of plume models to create a safer world for nuclear energy.

The Intelligence Community Postdoctoral Research Fellowship Program is funded by the Office of the Director of National Intelligence (ODNI) and managed by the Oak Ridge Institute for Science and Education (ORISE) under an agreement between the IC and the U.S. Department of Energy (DOE). ORISE is managed for DOE by ORAU.

From the ORISE Featurecast:

Using machine learning to model fallout plumes from CBRN incidents: A conversation with Nathan Boyle

Nathan Boyle, a former postdoctoral fellow in the Intelligence Community Postdoctoral Research Fellowship Program, has studied machine learning for predicting fallout from chemical biological radiological and nuclear weapons. In this episode of the ORISE Featurecast, Boyle discusses his research emphasis during his fellowship, where he is now, the value of mentorship and collaboration in the research process and so much more.

Listen Now!

Video Spotlight

ORISE Success Stories Video Profile

Hear from Dr. Nathan Boyle in his own words.