An open-source simulator built on Unreal Engine for drones, cars and other autonomous vehicles, with APIs for AI and computer vision research.
iamaisim/ProjectAirSim
An open-source, engine-independent simulator for drones, fixed-wing aircraft and ground vehicles, with optional Unreal Engine 5 rendering.
- Stars
- 938
- Forks
- 180
- Open issues
- 53
- Last commit
- 15 days ago
What is it?
Project AirSim is a simulation platform for building, testing and training autonomous systems such as drones, fixed-wing aircraft and unmanned ground vehicles. It is the continuation of Microsoft's AirSim, maintained by IAMAI Consulting Corp. after Microsoft discontinued the Project AirSim initiative.
The simulation core and client APIs can run in a lightweight, engine-independent runtime without Unreal Engine. When a 3D world, rendered cameras, LiDAR or radar are needed, the Unreal Engine 5 plugin provides the environment and rendering. Physics can come from the built-in Fast Physics, JSBSim for fixed-wing flight dynamics, or Simulink models.
Autonomy software connects through Python and C++ client libraries or a ROS 2 C++ bridge, and flight controllers such as PX4 and ArduPilot can run in software-in-the-loop or hardware-in-the-loop setups.
Key features
- 01Single and multi-vehicle scenes with configurable robots and sensors.
- 02Fast Physics, JSBSim flight dynamics and Simulink physics models.
- 03PX4 (SITL/HITL) and ArduPilot SITL flight-controller workflows.
- 04Python and C++ client libraries plus a ROS 2 C++ bridge.
- 05Cameras, LiDAR, radar, IMU, GPS and other sensors, with optional headless operation.
Use cases
- Developing and testing flight controllers or autonomy stacks before flight tests.
- Simulating drone swarms or mixed ground and aerial vehicle teams.
- Running headless simulations on cloud or CI machines without a GPU-rendered display.
- Migrating existing Microsoft AirSim workflows to a maintained codebase.
Quick start
- 1
Download a packaged environment from GitHub Releases
- 2
Follow the pre-built environment guide to launch it and control a vehicle with Python
Things to know
- Full rendering needs Unreal Engine 5.2, 5.7 or 5.8, and hardware requirements are driven by Unreal's rendering workload.
- Project AirSim Runtime does not provide cameras, LiDAR, radar or Unreal world meshes.
- Ubuntu 22.04 is the primary supported Linux distribution, and macOS support is limited to CI-tested builds.
In collections
Topics
FAQ
- What is ProjectAirSim?
- An open-source, engine-independent simulator for drones, fixed-wing aircraft and ground vehicles, with optional Unreal Engine 5 rendering. Project AirSim is a simulation platform for building, testing and training autonomous systems such as drones, fixed-wing aircraft and unmanned ground vehicles. It is the continuation of Microsoft's AirSim, maintained by IAMAI Consulting Corp. after Microsoft discontinued the Project AirSim initiative.
- Is ProjectAirSim free to use? What license does it have?
- ProjectAirSim is open source under the MIT license. Check the license terms before using it in commercial products.
- Is ProjectAirSim still maintained?
- Actively maintained. The last commit to ProjectAirSim was on Oct 23, 2026.
- What can I use ProjectAirSim for?
- Developing and testing flight controllers or autonomy stacks before flight tests. Simulating drone swarms or mixed ground and aerial vehicle teams. Running headless simulations on cloud or CI machines without a GPU-rendered display. Migrating existing Microsoft AirSim workflows to a maintained codebase.
- How do I install ProjectAirSim?
- Download a packaged environment from GitHub Releases Follow the pre-built environment guide to launch it and control a vehicle with Python
- Where can I get ProjectAirSim?
- The source code and installation instructions are on GitHub: https://github.com/iamaisim/ProjectAirSim
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A Unreal Engine-based simulator for drones, cars and other vehicles, extended by Cosys-Lab with new sensor models and APIs.
An open ROS2 framework to simulate and deploy perception-based drone swarms across PX4, ArduPilot and Betaflight autopilots.
Open-source robot simulator for modeling, programming and simulating robots, vehicles and mechanical systems.
A framework on NVIDIA Isaac Sim for simulating multirotor drones, with PX4 and ArduPilot autopilot integration and a Python control interface.
Lightweight 2.5D dynamics simulator for wheeled mobile robots and vehicles, usable standalone, via ROS 1/2, or from C++ and Python.