Artificial LIfe ENvironment (ALIEN) is an artificial life simulation tool based on a specialized 2D particle engine in CUDA for soft bodies and fluids. Each simulated body consists of a network of particles that can be upgraded with higher-level functions, ranging from pure information processing capabilities to physical equipment (such as sensors, muscles, weapons, constructors, etc.) whose executions are orchestrated by neural networks. The bodies can be thought of as agents or digital organisms operating in a common environment. Their blueprints can be stored in genomes and passed on to offspring.
The simulation code is written entirely in CUDA and optimized for large-scale real-time simulations with millions of particles. The development is driven by the desire to better understand the conditions for (pre-)biotic evolution and the growing complexity of biological systems. An important goal is to make the simulator user-friendly through a modern user interface, visually appealing rendering and a playful approach.
Please join our Discord server as a place for discussions, new developments and feedback around ALIEN and artificial life in general.
Demo video: Emerging Ecosystems | Winner of the ALIFE 2024 Virtual Creatures Competition
- Particles for simulating soft and rigid body mechanics, fluids, heat dissipation, damage, adhesion etc.
- Real-time user interactions with running simulations
- Simulation runs entirely on GPU via CUDA
- Rendering and post-processing via OpenGL using CUDA-OpenGL interoperability
physics2.mp4
- Multi-cellular organisms are simulated as particle networks
- Genetic system and cell by cell construction of offspring
- Neural networks for controlling higher-level functions (e.g. sensors and muscles)
- Various colors may be used to customize cell types according to own specifications
- Support for spatially varying simulation parameters
alife3.mp4
- Graph editor for manipulating every particle and connection
- Freehand and geometric drawing tools
- Genetic editor for designing customized organisms
- Mass-operations and (up/down) scaling functions
- Built-in simulation browser
- Download and upload simulation files
- Upvote simulations by giving stars
- A first attempt to answer: Feed your curiosity by watching evolution at work! As soon as self-replicating machines come into play and mutations are turned on, the simulation itself does everything.
- Perhaps the most honest answer: Fun! It is almost like a game with a pretty fast and realistic physics engine. You can make hundreds of thousands of machines accelerate and destroy with the mouse cursor. It feels like playing god in your own universe with your own rules. Different render styles and a visual editor offer fascinating insights into the events. There are a lot of videos on the YouTube channel for illustration.
- A more academic answer: A tool to tackle fundamental questions of how complexity or life-like structure may arise from simple components. How do entire ecosystems adapt to environmental changes and find a new equilibrium? How to find conditions that allow open-ended evolution?
- A tool for generative art: Evolution is a creative force that leads to ever new forms and behaviors.
A documentation for the previous major version, which introduces the reader to the simulator with tutorial-like articles, can be found at alien-project.gitbook.io/docs. Please notice that many of the information therein are no longer up to date. The latest version includes a brief documentation and user guidance in the program itself via help windows and tooltips.
Further information and artwork:
An Nvidia graphics card with compute capability 6.0 or higher is needed. Please check https://en.wikipedia.org/wiki/CUDA#GPUs_supported.
Installer for Windows: alien-installer.msi (updated: 2024-11-24)
In the case that the program crashes for an unknown reason, please refer to the troubleshooting section below.
The build process is mostly automated using the cross-platform CMake build system and the vcpkg package manager, which is included as a Git submodule.
To obtain the sources, please open a command prompt in a suitable directory (which should not contain whitespace characters) and enter the following command:
git clone --recursive https://github.com/chrxh/alien.git
Note: The --recursive parameter is necessary to check out the vcpkg submodule as well. Besides that, submodules are not normally updated by the standard git pull command. Instead, you need to write git pull --recurse-submodules.
Prerequisites: CUDA Toolkit 11.2+ and a toolchain for CMake (e.g. GCC 9.x+ or MSVC v142+).
Build steps:
mkdir build && cd build
cmake .. -DCMAKE_BUILD_TYPE=Release
cmake --build . --config Release -j8
If everything goes well, the ALIEN executable can be found under the build directory in ./alien or .\Release\alien.exe depending on the used toolchain and platform.
It is important to start ALIEN directly from the build folder, otherwise it will not find the resource folder.
There are reported build issues with (updated: 2024-06-22)
- GCC 12+ (version 11 should work)
- Visual Studio 17.10 (version 17.9 should work)
- CUDA 12.5 (version 12.4 should work)
This repository also contains a CLI for ALIEN. It can be used to run simulations without using a GUI. This is useful for performance measurements as well as for automatic execution and evaluation of simulations for different parameters. The CLI takes the simulation file, along with its parameters and the number of time steps, as input. It then provides the resulting simulation file and the statistics (as a CSV file) as output. For example,
.\cli.exe -i example.sim -o output.sim -t 1000
runs the simulation file example.sim for 1000 time steps.
Please make sure that:
- You have an NVIDIA graphics card with compute capability 6.0 or higher (for example GeForce 10 series).
- You have the latest NVIDIA graphics driver installed.
- The name of the installation directory (including the parent directories) should not contain non-English characters. If this is not fulfilled, please re-install ALIEN to a suitable directory. Do not move the files manually. If you use Windows, make also sure that you install ALIEN with a Windows user that contains no non-English characters. If this is not the case, a new Windows user could be created to solve this problem.
- ALIEN needs write access to its own directory. This should normally be the case.
- If you have multiple graphics cards, please check that your primary monitor is connected to the CUDA-powered card. ALIEN uses the same graphics card for computation as well as rendering and chooses the one with the highest compute capability.
- If you possess both integrated and dedicated graphics cards, please ensure that the alien-executable is configured to use your high-performance graphics card. On Windows you need to access the 'Graphics settings,' add 'alien.exe' to the list, click 'Options,' and choose 'High performance'.
If these conditions are not met, ALIEN may crash unexpectedly.
If the conditions are met and the error still occurs, please start ALIEN with the command line parameter -d, try to reproduce the error and then create a GitHub issue on https://github.com/chrxh/alien/issues where the log.txt is attached.
Contributions to the project are very welcome. The most convenient way is to communicate via GitHub Issues, Pull requests or the Discussion forum depending on the subject. For example, it could be
- Providing new content (simulation or genome files)
- Producing or sharing media files
- Reporting of bugs, wanted features, questions or feedback via GitHub Issues or in the Discussion forum.
- Pull requests for bug fixes, code cleanings, optimizations or minor tweaks. If you want to implement new features, refactorings or other major changes, please use the Discussion forum for consultation and coordination in advance.
- Extensions or corrections to the alien-docs. It has its own repository.
A short architectural overview of the source code can be found in the documentation.
ALIEN has been initiated, mainly developed and maintained by Christian Heinemann. Thanks to all the others who contributed to this repository:
The following external libraries are used:
- CUDA Toolkit
- Dear ImGui
- ImPlot
- ImFileDialog
- boost
- Glad
- GLFW
- glew
- stb
- cereal
- zlib
- zstr
- OpenSSL
- cpp-httplib
- googletest
- vcpkg
- WinReg
- CLI11
Free icons and icon font:
- IconFontCppHeaders
- Iconduck (Noto Emoji by Google, Apache License 2.0)
- Iconfinder (Bogdan Rosu Creative, CC BY 4.0)
- People icons created by Freepik - Flaticon (Flaticon license)
ALIEN is licensed under the BSD 3-Clause license.
