Sessions

GRID Session is a unified environment for robotics development that combines interactive coding, simulation, telemetry, and data management.

​

Overview

A Session is your development environment, where you can:

• Prototype code in Jupyter-style notebooks (left pane).
• View and interact with a real-time simulation (right pane).
• Monitor telemetry and logs in a dedicated panel (bottom-right).
• Leverage AI Copilot to generate actions or code from plain language.
• Store and retrieve data in the built-in file storage.

This integrated interface facilitates quick iteration: you write code in the left pane, immediately see your changes affect the simulation on the right, and track telemetry data as you go.

​

Layout

1. Notebook Panel (Left)

   • Write Python code in a Jupyter-like interface.
   • Execute cells to run code, instantiate robots or models, and visualize results in the simulation view.
   • Organize your workflow with Markdown cells for documentation and commentary.

2. Simulation Panel (Right)

   • Displays the simulation environment in which your code runs.
   • Interact with the simulation scene (e.g., position objects, change parameters).

3. Telemetry Tab (Bottom-Right)

   • Powered by Rerun, the Telemetry tab logs and visualizes robot states, sensor data, and results of AI models.
   • Helps you track performance metrics, movement paths, and any events or anomalies over time.

4. Copilot Tab (Bottom-Right)

   • Natural Language to Code conversion to help you prototype quickly.
   • Enter plain-language commands and Copilot will generate Python code tailored to your environment and simulation.
   • Seamless integration with your Notebook: copy/paste or insert Copilot-generated code directly into your active session.

5. Storage

   • Access saved data (images, logs, configuration files, etc.) generated during the session.
   • Download or upload files needed for your simulation or code.
   • Data persists across session restarts as long as you keep the session active.

​

Typical Workflow

1. Initialize Objects

   • In the left notebook panel, import the relevant robotics modules.
   • Create simulation objects, such as robots and sensors, by running AirGenCar(), OneFormer(), or other constructors.

2. Interact with the Simulation

   • Observe real-time changes in the simulation window when you execute your code.
   • Manipulate objects: move them around, apply commands, or adjust sensor inputs.
   • Ensure your setup reflects the scenario you want to prototype (e.g., suburban delivery routes, robotic arms in a warehouse, etc.).

3. Monitor Telemetry

   • As the simulation runs, switch to the Telemetry tab.
   • Inspect logs and visualize path trajectories or sensor data in real time.
   • Use this data for debugging or performance optimization.

4. Use Copilot

   • When you have a desired outcome in mind but aren’t sure how to implement it, open the Copilot.
   • Describe what you want in plain language (e.g., “Move the robot forward for 5 seconds”).
   • Integrate the auto-generated code snippet into your Notebook and execute it to see results.

5. Save & Download Data

   • As your session progresses, snapshots, logs, or analysis results will be saved in the Storage panel.
   • You can rename or structure files as needed, then download them for further offline analysis.

Step onto the GRID—where every idea takes shape and takes off.