Skip to main content
AirGen ships 26 vehicles. Each one is driven from Python by one of three classes, chosen by the vehicle’s physics model — you pick the robot by id when you start the session, and the notebook instantiates the matching class for you.

​
Aerial

Multirotors, flown through AirGenDrone on AirGen’s simpleflight model. Fixed-wing runs on the JSBSim flight-dynamics model and is also driven through AirGenDrone:

​
Ground

Wheeled and tracked vehicles, driven through AirGenCar on Unreal’s Chaos vehicle physics.

​
Legged

Quadrupeds, driven through AirGenQuad.

​
Camera-only

unitree_go2 and ghost_v60 also exist as physical robots GRID drives through grid-nexus-client. Because both the simulated and physical versions implement the same Robot interface, a skill you develop against the simulated Go2 runs against the real one.

Related topics

CLI referenceRemoteRobotThe Intelligence Grid for Physical AICommon FAQsYour first session