(734) 763-3867

Applications:
Energy Research, Robotics, Sensors and Sensor Networks
Methodologies:
Bayesian Methods, Deep Learning, Dynamical Models, Machine Learning, Predictive Modeling, Real-time Data Processing, Sparse Data Analysis, Statistical Inference, Time Series Analysis

Brendan Kochunas

Assistant Professor

Nuclear Engineering and Radiological Sciences

Dr. Kochunas’s research focus is on the next generation of numerical methods and parallel algorithms for high fidelity computational reactor physics and how to leverage these capabilities to develop digital twins. His group’s areas of expertise include neutron transport, nuclide transmutation, multi-physics, parallel programming, and HPC architectures. The Nuclear Reactor Analysis and Methods (NURAM) group is also developing techniques that integrate data-driven methods with conventional approaches in numerical analysis to produce “hybrid models” for accurate, real-time modeling applications. This is embodied by his recent efforts to combine high-fidelity simulation results simulation models in virtual reality through the Virtual Ford Nuclear Reactor.

Relationship of concepts for the Digital Model, Digital Shadow, Digital Twin, and the Physical Asset using images and models of the Ford Nuclear Reactor as an example. Large arrows represent automated information exchange and small arrows represent manual data exchange.