Aravind Sunil Kumar

Building AI agents and making surgical robots more accessible

prof_pic.jpg

I’m an engineer at NVIDIA. My research focuses on 3D scene understanding for surgical environments - spanning surgical robotics, neural rendering, tissue deformation modeling, and spatially-aware perception for image-guided procedures.

I also work on AI agents, building systems for agent-based diagnostics that integrate structured reasoning and multimodal signals.

I have contributed extensively to open-source surgical robotics, with contributions to the da Vinci Research Kit (dVRK) ecosystem and ROS 2 - 3D Slicer medical imaging integrations.

I have collaborated with Anton Deguet, Axel Krieger, and Simon Leonard on improving ROS 2 and 3D Slicer integration, including work on SlicerROS2, ROS 2 pipelines, and surgical simulation workflows. Earlier, I worked with Dr. Soumyadipta Acharya on vectorCam, a computer-vision project that went on to be featured on Bill Gates’ youtube.

Academically, I completed my Master’s in Computer Science at Johns Hopkins University, where I worked within the medical robotics community, and my B.Tech in Mechanical Engineering at IIT Madras.

Current Technical Interests

  • AI agents for diagnostics and structured reasoning
  • 3D scene understanding for surgical environments
  • Deformable tissue representations and neural rendering
  • Surgical robotics research infrastructure (ROS 2 / 3D Slicer / Isaac Sim)
  • Video-based action learning for vision-language-action models

I enjoy collaborating on problems at the intersection of medical robotics, 3D scene understanding, and AI systems, and I’m always happy to connect with researchers or engineers working in these areas.

Updates

  • SlicerROS2 — Introductory Talk at ROS2 Meetup Bogota, 2025 A technical introduction to bridging 3D Slicer and ROS 2 for robotic surgery research. Watch on YouTube

  • SlicerROS2 v1.0 Release Documentation and tooling for integrating ROS 2 pipelines into 3D Slicer. V1 Release Documentation

  • VectorCAM featured on Bill Gates’ Youtube Channel vectorCam featured in a Bill Gates video on scalable global health diagnostics. Watch on YouTube

  • AI Systems for Diagnostic Analysis - Talk at RECon, Dublin 2025 A talk on AI agent architectures for diagnostic reasoning — integrating structured models with LLMs for diagnosis. Watch on YouTube

  • MedMachina — Medical Robots Directory https://medmachina.github.io/

selected publications

  1. IEEE T-MRB
    SlicerROS2: A Research and Development Module for Image-Guided Robotic Interventions
    Laura Connolly, Aravind S. Kumar, Kapi Ketan Mehta, and 9 more authors
    IEEE Transactions on Medical Robotics and Bionics, Nov 2024
  2. Sci. Rep.
    Towards transforming malaria vector surveillance using VectorBrain: a novel convolutional neural network for mosquito species, sex, and abdomen status identifications
    Deming Li, Shruti Hegde, Aravind Sunil Kumar, and 8 more authors
    Scientific Reports, Nov 2024