Data Science Institute

Data Scientist Spotlight

Mason Sage

Staff Robotics Engineer and Principal Investigator

What is your job title/position? Staff Robotics Engineer and Principal Investigator

Which directorate/division do you support? I sit in the Materials Engineering Division of Engineering, but I support a wide range of projects across Physical & Life Sciences and Engineering.

When did you come to the Lab? June 2022

What did you study in your path to this career? I studied mechatronics in trade school, then mechanical engineering and computer science at university.

What project(s) are you currently working on? I am mainly working on Project APEX (Autonomous Alloy Prediction and EXploration), a Laboratory Directed Research and Development Exploratory Research project and LLNL’s first self-driving lab for designing new alloys. We combine robotics with machine learning to design, build, prepare, and characterize novel metals.

How would you characterize your career motivation? I am happiest when I am trying something new. In almost a decade of work, I have never had two years that were the same. Whether it is tackling technical challenges or stepping into more of a management role, I appreciate the culture here and the flexibility to keep growing and exploring new paths in my career.

ParaView-MCP levels the playing field for complex scientific visualization

Visualizing the structures, forces, and processes involved in scientific research projects is essential for clearly conveying complex aspects of experiments. However, outside of designing, running, and analyzing their experiments, few scientists have enough time to commit to learning the software tools used to create scientific visualizations.

LLNL researchers Shusen Liu, Haichao Miao, and Peer-Timo Bremer from LLNL’s Center for Applied Scientific Computing (CASC) set out to enable scientists to make use of a visualization tool without extensive training and allow for direct, expert input into the visualization design process. They focused on ParaView, a premier, open-source application for scientific visualization used across the National Laboratories. The outcome of their study, called ParaView-MCP, empowers users to interact with the application through natural-language and visual inputs instead of the typical graphic user interface (GUI), which can appear daunting for novice users. Their work helps lower the barrier of entry to using ParaView’s capabilities while empowering users by autonomously conducting analysis of complex datasets via AI-driven decision support.