Model-Based Deep Learning Accelerated Cosmic Ray Muon Tomography for Real-Time Automated Monitoring of Advanced Microreactors - College of Engineering - Purdue University

Project Description

This project will develop and validate a model-based methodology for non-invasive monitoring of microreactor cores using cosmic ray muon tomography (muon CT). Microreactors, envisioned for remote or energy-limited communities, require robust safeguards and operational monitoring solutions that do not interfere with reactor operations. Cosmic ray muons, naturally occurring high-energy particles, provide a unique opportunity to image dense materials such as nuclear fuel in real time, not feasible with conventional x-ray or neutron imaging techniques.
The post-doctoral fellow will perform high-fidelity Geant4 simulations of microreactor geometries, modeling the interactions between cosmic ray muons and reactor materials to generate synthetic detector datasets. These datasets will be used to develop three-dimensional model-based iterative reconstruction algorithms capable of resolving the spatial distribution of fuel assemblies and detecting anomalies. To accelerate identification of missing or displaced fuel, the project will integrate deep learning approaches, enabling rapid and automated analysis of large muon tomography datasets.
This project combines nuclear engineering, particle physics, computational imaging, and artificial intelligence. By providing passive monitoring of microreactor cores, it will support fuel verification, early anomaly detection, and nuclear nonproliferation.

Start Date

August 2026

Postdoc Qualifications

Co-advisors

Bibliography