Richard Vasques

Assistant Professor of Nuclear Engineering

Machine Learning in Radiation Detection and Diagnostics


Accurate neutron energy spectrum information is important to both national security and personal safety. Reconstructing a neutron energy spectrum from detector responses is a heavily researched problem, given the role of neutron energy in determining radiation dose. This project investigates the passive neutron spectrometer (PNS), a detection system developed in collaboration with Lawrence Livermore National Laboratory, for use in energy spectrum unfolding primarily in the event of a criticality accident. The PNS provides a passive detection method through 55 thermoluminescent dosimeters or gold foils contained within a single polyethylene sphere. The unfolded spectrum is used to calculate the dose a person would receive in the corresponding neutron field. Three unfolding algorithms are employed, including a neural network developed in our group. 
The broader research program extends this framework to additional diagnostic applications, including reactor facility characterization and detector calibration, with ongoing work exploring the same computational unfolding methods in detector configurations beyond the original PNS design.
Doctoral Dissertation advised on the subject:

Publications


[J25] Introduction, investigation, and experimental validation of a novel Passive Neutron Spectrometer


Zachary T. Condon, Daniel Siefman, Paul Maggi, Paige Witter, Richard Vasques

Nuclear Science and Engineering, vol. 199, 2025 Sep, pp. 1546--1562


[P34] Using machine learning to unfold neutron spectra wth a passive neutron spectrometer


Zachary Condon, Daniel Siefman, Richard Vasques

Proceedings of International Conference on Mathematics & Computational Methods Applied to Nuclear Science & Engineering, Niagara Falls, Canada, 2023 Aug