Agentic Workflows in Isaac for Healthcare#

From prompt to policy

Prompt Your Way to a Healthcare Robotics Simulation

Isaac for Healthcare supplies the simulation, data, and policy building blocks for robotic autonomy. In this course, you use agentic workflows to create a healthcare task, collect demonstrations, train a policy, and validate it in Isaac Sim.

This hands-on course teaches you to build, simulate, and extend healthcare-robotics workflows with NVIDIA Isaac for Healthcare, the unified IsaacLab-Arena simulator, GR00T, and openpi policies. You’ll work entirely in simulation, following the same record → prepare → train → deploy loop the platform’s agentic skills automate.

The course is organized as self-paced sections. Work through them, then extend the platform with your own environments and skills.

Compute Requirements#

This is a GPU course. Before you begin, use a local or remote machine with:

  • Ubuntu 22.04 or 24.04 on x86_64 or aarch64.

  • An NVIDIA GPU with Linux driver 580.65.06 or newer for Isaac Sim 5.1.0 and CUDA 13.

  • About 30 GB of free disk space.

  • git and uv available on PATH.

  • Network access to GitHub, PyPI, NVIDIA PyPI, and PyTorch CUDA wheels.

Cosmos visual expansion is optional. It also needs Docker with GPU support and an HF_TOKEN or HUGGING_FACE_HUB_TOKEN after you accept the NVIDIA Cosmos model license.

Stay in the CLI With the Tutor Skill#

If you prefer to stay in your terminal the whole time, load the Physical AI Tutor skill into your coding agent before you begin. It turns this course into an interactive CLI tutor: it adapts to your experience level, explains one step at a time, pauses before hands-on actions, and helps diagnose errors from the same shell session.

Tutor me through Agentic Workflows in Isaac for Healthcare from the beginning.
Use the course instructions at https://docs.nvidia.com/learning/physical-ai/getting-started-with-isaac-for-healthcare/latest/agentic-workflows.
Keep me in the CLI, adapt to my level, and pause before each hands-on step.
The Isaac for Healthcare agentic end-to-end pipeline

The end-to-end agentic pipeline runs in four phases: prepare → build dataset → inspect & convert → train & verify.#

What You’ll Do#

Get Set Up

Install the workflow, list the supported environments, and run a zero-action smoke test in Setup and Environments.

Build a Dataset

Collect teleop demos and prepare them into a LeRobot dataset.

Train and Deploy

Fine-tune a GR00T policy and deploy it in simulation.

Extend the Platform

Build your own workflow and add your own skills.