MRS FlightForge Simulator

Requirements

• A dedicated NVidia GPU (at least MX450, RTX is recommended) or a comparable AMD GPU
• 3 GB of HDD space

The Maps

Map Name		Map Name
Valley		Cave
Forest		Erding Airbase
Warehouse		Infinite forest

Controls

• m/ESC - show menu for switching graphics and worlds

Installation

Linux

1. Install the MRS UAV System
2. Install the MRS UAV System FlightForge endpoint

sudo apt install ros-noetic-mrs-uav-unreal-simulation

3. Download the and unpack MRS FlightForge simulator
4. Launch the FlightForge Simulator by ./mrs_flight_forge.sh. Alternatively, start it in a headless mode via ./mrs_flight_forge.sh -RenderOffscreen.
5. Start the MRS UAV System FlightForge endpoint

roscd mrs_uav_unreal_simulation
./tmux/one_drone/start.sh

Windows

Requirements:

• Windows 11 + updates
• WSL 2.0

1. Install Ubuntu 20.04 into the WSL 2.0 using the Microsoft Store
2. Install the MRS UAV System into WSL 2.0
3. Create .wslconfig file in C:/Users/<User>/.wslconfig
4. Place the following content into .wslconfig

[wsl2]
firewall=false
networkingMode=mirrored
debugConsole=false
[experimental]
hostAddressLoopback=true

5. Restart the WSL by issuing wsl --shutdown into a comand line.
6. Install the MRS UAV System FlightForge endpoint

sudo apt install ros-noetic-mrs-uav-unreal-simulation

7. Download and unpack the FlightForge Simulator
8. Start the FlightForge Simulator
9. Start the MRS UAV System FlightForge endpoint

roscd mrs_uav_unreal_simulation
./tmux/one_drone/start.sh

10. The first start might require hitting "CTRL+C" in the roscore tab of the tmux session. The roscore is always stuck for the first time after rebooting the computer.

Citing this work

If you use this simulator in your research, please cite the following paper:

  title   = {FlightForge: Advancing UAV Research with Procedural Generation of High-Fidelity Simulation and Integrated Autonomy},
  author  = {David Čapek and Jan Hrnčíř and Tomáš Báča and Jakub Jirkal and Vojtěch Vonásek and Robert Pěnička and Martin Saska},
  year    = {2025},
  journal = {arXiv preprint arXiv: 2502.05038},
  url     = {https://arxiv.org/abs/2502.05038v1},
  pdf     = {https://arxiv.org/pdf/2502.05038.pdf}
}

Planned features

• Adding back a grayscale normalized depth image
• possible coloring of the depth pointcloud using the segmented RGB image