Getting Started

Once a session is launched and running, here are a few steps to get started with GRID!

1. Confirm the simulation is running and streaming correctly. Initially, you will see a "Connecting..." message in the simulation window which will then be replaced by the stream.

2. Run the very first cell in the notebook to initialize the simulation "client", and the second cell (if present) to initialize any AI models that were loaded.

3. Let's try to control the vehicle now. If using the drone, add a cell to the notebook and write the following commands in the cell:

airgen_drone_0.client.takeoffAsync()

Once you press the Run button, you should see the drone take off and hover in the air. For more information on how to control the drone through various methods, please refer to the AirGen documentation.

4. Now let's try to capture an image and visualize it in the telemetry tab. Add another cell, and write the following code in it:

import airgen, rerun as rr

image_data = airgen_drone_0.client.getImages("front_center", [airgen.ImageType.Scene])
rgb_image, camera_pose = image_data[0]
rr.log("image", rr.Image(rgb_image))

5. Once the above cell is run, you should see the image appear in the telemetry tab.

6. Switch to the Terminal tab, and try chatting with the LLM. For example, you could enter "Land the drone" at the prompt, after which the LLM will propose an execution plan as a step-by-step process. After that, clicking on execute will make the LLM generate the appropriate code (which will appear in the notebook automatically) and the results can be seen in the simulation window.