Flyappy Autonomy Test Game

This repository contains the coding test for the Flyability Autonomy team.

Game Description

Flyappy is in trouble again! This time it went into space and landed in an asteroid belt. If Flyappy collides with the asteroids it would be fatal. Luckily Flyappy remembered his laser scanner that provides distance measurements. It will give you its velocity and the laserscans in return for an acceleration input. Flyappy only asks for 60 seconds of your guidance. Help Flyappy go through as many asteroid lines as possible before the time runs out!

Getting Started

Clone the repository:

git clone https://github.com/Flyability/flyappy_autonomy_test_public.git
cd flyappy_autonomy_test_public

Setup the running environment

This game has been tested with Ubuntu 24.04 running ROS2 Jazzy and Python 3.12.

There are two recommended options for running the game. Either use Ubuntu 24.04 and add the necessary packages to your system to compile and run the game. Or on any Linux distribution, use Distrobox to run an Ubuntu 24.04 distribution inside your terminal.

Option 1 - Using your system

If you already have Ubuntu 24.04 on your system, great. If not, you can either install it on your machine (dual-boot or full installation) or boot from an USB flash drive. You can follow Ubuntu tutorial: Install Ubuntu desktop.

If your system is running Windows 11, you can maybe try Ubuntu tutorial: Install Ubuntu on WSL2 on Windows 11 with GUI support.

Then, make sure ROS2 Jazzy is installed by following the ROS install guide.

Option 2 - Using Distrobox

First, install Docker.

Build an image with Ubuntu 24.04 and ROS 2 Jazzy from the ./Dockerfile:

docker build . -f Dockerfile -t flyappy

Then, install a recent version of Distrobox:

git -C /tmp clone https://github.com/89luca89/distrobox.git
cd /tmp/distrobox
git checkout 1.8.0
sudo ./install

Create a Distrobox image from the previously built docker image. The command takes around 2-3 minutes to complete and will ask to type a new password (twice):

distrobox-create --name flyappy --image flyappy --home ~/.flyappy/ && distrobox enter flyappy -- bash -cl true

Now, you can enter the Distrobox whenever you need with:

distrobox enter flyappy --clean-path

Inside the image, you can install the python packages, run the game, build the C++ code, and run your code. You can still edit the code from your main distribution.

Setting up python environment

Make sure python venv is installed:

sudo apt install python3.12-venv

Create a python virtual environment (with ROS2 packages available):

source /opt/ros/jazzy/setup.bash
python3 -m venv .venv --system-site-packages

Activate the python virtual environment:

source .venv/bin/activate

Install the main game:

pip install ./flyappy_main_game

Run The Game

To run the game, launch (inside the python virtual environment):

flyappy_main_game

A GUI will become visible with the game start screen.

You can then add velocity in a wanted direction by pressing the arrow keys ←↑↓→.

Notice that it is not possible to go backwards and if Flyappy hits an obstacle the game stops.

Automate Flyappy

Now that we have gotten familiar with the game, we want to control Flyappy autonomously. To do this, a Python and a C++ template have been provided.

For now the autonomy code does not do anything other than printing out some laser ray and end game information. To start the game, press any arrow key.

Python

Install the package (inside the python virtual environment):

pip install ./flyappy_autonomy_code_py

To run the automation code:

flyappy_autonomy_code_node

Tip

You can install in editable mode with the pip option -e to avoid installing every time you do a change in the Python code.

C++

Compile the code:

cd flyappy_autonomy_code_cpp
source /opt/ros/jazzy/setup.bash
cmake --preset release
cmake --build --preset release

To run the automation code:

./build/release/flyappy_autonomy_code_node

Tip

Unit tests can be run with:

ctest --preset release

You can also compile in Debug mode with --preset debug

Modifying the code

The templates are located in the flyappy_autonomy_code_cpp/ and flyappy_autonomy_code_py/ folders. Be aware that you are not meant to change the files in flyappy_main_game/.

For using python, modify (or add) any files in the src/, tests/ folders.

For using C++, modify (or add) any files in the src/, tests/ folders and, if needed, the CMakeLists.txt.

Take your pick.

To get the state of Flyappy, its velocity and laserscans data are published on 2 ROS topics. An acceleration command can be given on a ROS topic for actuating Flyappy.

The callbacks and publisher are provided in the code.

Handing In

To hand in your game solution, please send your flyappy_autonomy_test_public repository in a ZIP file by email.

Other Information

I hope you will have fun solving this little game. If you have any questions or need other game information either write us or look around in the flyappy_main_game folder. Here is some other helpful information for solving the task.

• Scaling: 1 pixel = 0.01 meter
• Game and sensor update rates: 30 fps
• The velocity measurement is noise free
• Max acceleration x: 3.0 m/s^2
• Max acceleration y: 35.0 m/s^2
• Axis convention: x →, y ↑
• LaserScan message definition

Value	Unit	Topic
Velocity	m/s	/flyappy_vel
Acceleration	m/s^2	/flyappy_acc
LaserScan	Radians, meters	/flyappy_laser_scan
GameEnded	No unit	/flyappy_game_ended