Vic Castillo

About Vic Castillo

Scientist and former group leader in the Computational Engineering Division at Lawrence Livermore National Laboratory with a background in Computational Physics, Machine Learning, and System Dynamics with over 30 years of experience in industry and government research. He has been applying neural networks and other machine learning methods to solve manufacturing and engineering problems since the early 1990’s. He has a patent for using neural networks to control blow-molding systems with high post-consumer material for the Clorox bleach bottle. Principal investigator on over a dozen HPC4Mfg projects, including glass manufacturing modeling and optimization projects with PPG Flat Glass, Vitro Plate Glass, and Guardian Glass; Aluminum manufacturing projects with Alcoa and Arconic; Steel manufacturing and forming with ArcelorMittal, Cleveland Cliffs, AK Steel, and Machina Labs; and gas turbine designer VAST Power. Dr. Castillo is also leading a HPC team to develop novel sampling methods for active learning on the HPC environment. One of the principal investigators for the FY2013 Laboratory Directed Strategic Initiative Accelerated Certification of Additively Manufactured Metals which optimized Concept Laser Series 5 laser powder-bed metal additive manufacturing systems for NNSA applications. Dr. Castillo has a B.S. in Physics and M.S. in Computational Physics from San Jose State University and a Ph.D. in Applied Engineering from the University of California at Davis. Additionally, Vic contributes substantially to STEM outreach programs which seek to expose teachers and students to the value of computer modeling, AI, and robotics. He was honored with the National 2013 Community Service award from Great Minds in STEM and 2017 LLNL Director’s Lifetime Achievement Award for STEM outreach and diversity engagement.

Experience

1. Senior Research Scientist

   Lawrence Livermore National Laboratory

   Dec 1998 — Present · Livermore, CA, US

   Developing methods for Scientific Machine Learning (SciML):Reduced-Order Models from autoencoder to develop fast-running surrogate models of multi-phase flows; Machine-Learning workflow tools for High-Performance Computing; Active-Learning Design of Experiments for controlling experiments

Education

• University of California, Davis

  PhD, Engineering Applied Science

  1994 — 1999

• San José State University

  M.S. Computational Physics, Dynamical systems, Neural Networks, Pattern Formation

  1988 — 1992

• RAND School of Public Policy

  LLNL collaboration, Serious Gaming

  2017 — 2017

Skills

• Analysis
• Physics
• Applied Mathematics
• R
• Artificial Neural Networks
• Fortran
• Community Outreach
• Robotics
• Simulation
• Algorithms
• Discrete Event Simulation
• Mpi
• Computational Physics
• Artificial Intelligence
• Modeling
• Data Mining
• Science
• Mathematical Modeling
• Ros
• Python
• Speech Recognition
• Research
• Applied Physics
• Mathematica
• Natural Language Processing
• Monte Carlo Simulation
• Numerical Analysis
• Optimization
• Educational Technology
• Parallel Computing
• Operations Research
• Machine Learning
• Statistical Modeling
• Scientific Computing
• Simulations
• Fluid Dynamics
• High Performance Computing
• Image Processing
• Statistics