Arun Kothari

Responsibilities:

• Built a Unity-based AR/VR simulation for Navy MRO using Potree point-cloud rendering and Mirror networking; cut inspection setup time 40%.
• Set up Perforce + SQL pipeline to manage ship assets, annotations, and bounding boxes; reduced integration conflicts 70%.
• Added tooling: point-cloud editing, clash detection, and annotation system to speed up maintenance scenario iteration.
• Implemented ICP + RANSAC point-cloud registration for accurate multi-sensor ship alignment in AR/VR.
• Developing neural-network control system to replace heuristic UMV automation for more robust autonomous maintenance.

Responsibilities:

• Built and trained a U-Net deep learning model to segment and track oxygen atom movement in CeO₂ microscopy data, cutting manual annotation 60%.
• Applied deep learning image segmentation to high-resolution microscopy, improving consistency and accuracy of atomic-scale feature extraction.
• Developed an interactive React + p5.js web platform to visualize nanoparticle behavior for STEM education outreach.
• Delivered weekly technical briefings on model performance and experimental findings to guide research direction and collaboration.