AirGen Overview

​

Key Features

• Unreal Engine 5.7: Photorealistic rendering for perception workloads where image fidelity is the point.
• Aerial, ground, and legged robots: 26 vehicles across multirotors, fixed-wing, wheeled and tracked platforms, and quadrupeds — see Robots.
• Large outdoor environments: Substations, oil rigs, forests, warehouses, and planetary surfaces — see Scenes.
• Geographically anchored: Anchor the world origin to real latitude and longitude so GPS output matches a real site.
• Controllable conditions: Weather, wind, and a celestial clock that drives sun position from a wall-clock datetime.
• One robot interface: AirGen robots implement the same Robot interface as physical machines, so getImage() and moveByVelocity() mean the same thing in simulation and on hardware.

​

Start a session

GRID # session start sim

from grid_sim_client.airgen.robot import AirGenDrone

drone = AirGenDrone()
drone.takeoff()
image = drone.getImage("front")

​

Where next

• Robots — the vehicle catalog and which class drives each
• Scenes — the environments you can launch into
• Sensors — cameras, lidar, IMU, GPS, and more
• Session configuration — weather, time of day, and camera setup
• AirGen robots — the Python API reference