top of page

Rube Williams, Ph.D.

Research Interests:

  • Safe human-like rapid-moving autonomous vehicles

  • Artificial general intelligence

  • Safe super-intelligence

AI Related Publications and Patents

  1. Rube Williams, (2020-08-15), Better Than Human: Contextualized Automated Vehicle Intelligent Control, [Medium.com]

  2. Rube Williams (2021-08-27), Hello—Automatons!, [Medium.com]

  3. Rube Williams, Intelligent Two-Phase Flow Phenomena Sensor for Enhanced Thermal Management Control, NASA SBIR Final Report, Contract: 80NSSC19C0585, Feb, 2020 (no link) (summary) (patent)

  4. Rube B. Williams, Restricted Complexity Framework for Nonlinear Adaptive Control in Complex Systems, Space Technology and Applications international Forum, Albuquerque, NM, February 8-11, 2004; (article)

  5. R. Williams and A. Parlos, Adaptive State Filtering for Space Shuttle Main Engine Turbine Health Monitoring, Journal of Spacecraft and Rockets, Vol. 40, No. 1, pp. 101-109, 2003.; (article)

  6. R. B. Williams, Jr., Adaptive State Filtering with Application to Reusable Rocket Engines, Ph.D. Dissertation, Texas A&M University, 1997.(dissertation)

  7. Williams, R.B., Stratos Perception LLC, 2022. Systems and methods for monitoring and controlling a multi-phase fluid flow. U.S. Patent 11,341,657; (patent)

  8. Williams, R.B., Stratos Perception LLC, 2023. Systems and methods for Estimating Parameters of a Nonlinear Dynamic System. Patent Pending. (no link)

Software Skills and Academic Research

Languages/API/Tools/Libraries/IDE’s/Cloud: C/C++, NodeJS, Python, PyTorch, TensorFlow/Keras, Objective C, Swift, Javascript, SAS, Git, REST, XCode, MATLAB, Simulink, MCNPX, NumPy, Pandas, Scikit-learn, Docker, LabView, AWS, Jupyter, Atom

 

Graduate Studies @ Texas A&M University, College Station, TX

1992 – 1997 NASA Graduate Research Fellow, Nuclear Engineering

  • Ph.D. Nuclear Engineering, May 1997, Texas A&M University

  • Dissertation: Adaptive State Filtering with Application to Reusable Rocket Engines

  • Focus: Recurrent Neural Networks, Damage-Mitigating Control, Fault-Tolerant Control

Contributing Profession Experience

2021 – Present, Trice Imaging, San Diego, California (Remote)

  • Company Mission: Medical ultrasound imaging support

  • Position: Director, Machine Learning

 

AI Contributions:

  • Developed 3-year strategic plan for technology development

  • Developed machine vison system for interpreting maternal fetal medicine ultrasound images

  • Designed new productivity reporting to incorporating analytics from AI concerning sonography operations

  • Developed automation for MLOps, including training data curation and version control

  • Developed product demonstrators supporting new initiatives, including GPT co-pilot

  • Recruited and managed sonographer team to support science-based training data annotations

 

2018 – Present, Stratos Perception, LLC, Houston, Texas

  • Company Mission: Develop artificial intelligence solutions and software that increase productivity and reliability in aerospace systems

  • Position: Principal Investigator, Deep Learning, Software Developer

 

AI Contributions:

  • NASA SBIR Phase 1 Contract. ($110K award) Developed a machine vision process incorporating convolutional neural networks to perform multiphase flow management. (C/C++, Python, CUDA); Completed on time; Patent US 11,341,657 (2022)

    • Proposed, designed, and developed solution

    • Features YOLOv3, OpenCV, C/C++, Python, Supervised Learning

    • Performs real-time thermo-fluid dynamic analyses on two-phase channel flow based on machine-vision observations of the flow field at 1000 fps

  • Developed novel inferential sensor for estimating dynamical states and any number of time varying parameters, thereby providing first-principals observability to unanticipated failure modes; and also providing many advanced autonomous capabilities to machines. Patent Pending

2017-2019, Sunnova Energy Corporation, Houston, TX

  • Company Mission: To install and finance rooftop solar systems and battery storage systems for residential energy customers.

  • Position: Full Stack / Mobile Developer

AI Contributions:

  • Automated of the quality assurance inspection work for solar system installs, which were accomplished by engineers directly reviewing photographs of installed system components. Developed a deep learning system featuring a convolutional neural network utilizing object detection to automate part of the review of the photos to identify equipment installs and to extract serial numbers to a database.  Features YOLO, OpenCV, C/C++, Python, Supervised Learning, AWS (S3 and SQS), PostgreSQL

 

  • Modeled National Solar Radiation Data using LSTM network to supplement the published data, which included from 8%-30% error in the NOAA solar radiation data attributed to a local weather station. The objective was to effectively generalize the weather station data over the local regions of the US and more consistently predict the solar radiation, providing a shadow system to support better financial predictions. Python, TensorFlow, Keras, Jupyter, Supervised Learning

2000 – 2004   Los Alamos National Laboratory, Los Alamos, NM,  Technical Staff Member, Nuclear Systems Design and Risk Analysis, Decision Applications Division

AI Contributions:

  • Proposed a method utilizing neural networks to interpret human intent from the acceleration field of people wearing accelerometers  

bottom of page