Visualizations
Galaxy Evolution Visualized
My work focuses on using Galacticus to simulate the evolution of dark matter halos and our galaxies. However, sometimes it is hard to imagine what these simulations really look like. Here, you can watch and explore as a simulated galaxy grows and evolves over the age of the Universe. Try toggling "Galaxies only" to see what it would look like if we were to see it through a telescope, or choose "Dark matter only" to visualize the invisible structures that hold it all together. You can pause the simulation and hover over components to learn more!
Drag to rotate, scroll to zoom. Data is a static export of a single Galacticus tree from my dark-matter substructure simulations. Top 50 subhalos by infall mass are shown. This visualization was made using the help of AI coding agents.
Finding Defects in a 3D Printed Lattice Using AI Agents
Lattice defects
Drag to rotate, scroll or pinch to zoom, click a defect for details.
Struts and junctions of a 3D printed octet-truss lattice, reconstructed from its CT scan. Colored elements are the defects flagged by the analysis pipeline. This visualization was made using the help of AI coding agents.
Defects found in a CT scan of a 3D-printed titanium lattice. Our approach combines a traditional computer vision pipeline with an AI agent that dynamically classifies defect types. First, we find defects in the lattice using segmentation (converting the grayscale image into a black-and-white image, using a threshold to separate material from empty voxels) combined with simple "pixel checks" to find missing and partially missing (broken) struts and junctions with greater than 95% accuracy. Then our agent classifies the defects into two categories, "stochastic" and "systematic," and reasons about what may have caused each defect.