Downloads

- Deathmatch files zip

- Deathmatch tutorial pdf

Description

The Framsticks Deathmatch is an experiment which allows teachers to organize a competition, in which students compete by designing, creating, and evolving creatures to participate in the Deathmatch.

During the Deathmatch, these creatures must survive by eating food, they can fight and kill creatures of other teams, and finally die. The creators of the creatures that force all other teams to go extinct, and are the last surviving creatures, are the winners!

The Deathmatch is developed as educational platform to learn in artificial life and Framsticks by project-based and competitve learning.

Related articles

W. de Back, M, Komosinski, Deathmatch Tutorial (english), 2004.

Comprehensive tutorial for everyone interested in organising or participating in the Framsticks Deathmatch.

R. Veldkamp, W. de Back, Fight for your a-life (english), in: Software 2.0 Extra!, February 2005.

Article on the benefits of artificial life to game programming above game-AI. Features the Deathmatch as a way to get to hands-on experience with many alife-techniques in a game-like environment.
Published in a Poland-based software developers magazine, in Polish, German, French and Spanish.

W. de Back, Framsticks (dutch), in: de Connectie, Juli 2005.

General introduction to artificial life, with some special attention for artificial life in education. Features Framsticks as a platform for alife education, and the Deathmatch as an fun and didactive competition for students in AI.
Published in a Dutch periodical for AI students, teachers and researchers.

Artificial ecosystem

Downloads

- Ecosystem files zip

Note: requires Framsticks v2.10.10
Includes Framsticks theater show

Description

The artificial ecosystem simulates two populations of flying creatures, prey and predators, that interact with eachother and with their environment. This ecosystem implements a simple foodchain in which prey eat food and predators eat prey.
Due to these coupled struggle for existence, the sizes of these populations are related to each other. They exhibit a population dynamics that is described in the theoretical ecological models of Lotka-Volterra, including logistic growth and type 2 predator functional response (predator saturation).

This model is part of an ongoing project in which we study population dynamics, the interaction between population dynamics and evolutionary dynamics, competitive co-evolution.

Population sizes, population genetics and life histories are written to logfiles in scripts_output during experiments. These can be used for extensive analysis.

Related articles

W. de Back, Framsticks (dutch), in: de Connectie, Juli 2005.

General introduction to artificial life, with some special attention for artificial life in education. Briefly features virtual life ecosystem experiments as a bridge between evolutionary robotics and alife simulations like Tierra.
Published in a Dutch periodical for AI students, teachers and researchers.

W. de Back, Introduction to Endogenous evolution, Internal lab report, March 2004.

Simple introduction to the difference between conventional evolutionary algorithms and endogenous evolution. It is argued that genuine evolutionary self-organisation implies releasing the control and guidance of evolutionary search that is traditionally laid down in the fitness function. Instead of the traditional reliance of selection on fitness, endogenous fitness is based on natural selection (the other way around).

Downloads

- Khepera zip

Note: clear the network weights before evolving your own behaviour

Description

Many experiments in evolutionary robotics (ER) use Khepera robots. Therefore, a Khepera-like creature is developed in Framsticks to allow replication of projects reported in literature.
This (simulated) robot forms a platform that offers a wide variety of possibilities for experiments in evolution of behavioral tasks, neural networks and evolutionary learning mechanisms.
The creature is modelled in Framsticks using the f1 genotype encoding like X(X,X,X,X,X,X,X). On two opposing sides, two 'wheel neurons' are attached. The activation of the left wheel causes the creature to rotate around the right wheel (and visa versa). Sensors can be arranged on the endings of the sticks.

The particular khepera that can be downloaded here was evolved (optimised neural network weights) to find food.
Initially, the robot was situated in a world with 'teleport' boundary and evolved to adapt to this environment. This resulted in two kinds of behavior that was equally fit: (1) drive around like crazy and (2) go towards food. The first strategy fully depends on teleport, and the second doesn't.
Subsequently, the selection criteria is changed by changing the world boundary to 'fence' (while the fitness function stays the same!). This creates an advantage for the second strategy.

The interesting aspect of this particular robot is that it is capable of coping in an environment closed by a fence, while it is completely unable to sense any boundary.

Related literature

S. Nolfi, D. Floreano, Evolutionary Robotics: the Biology, Intelligence, and Technology of Self-organizing Machines, MIT Press, 2000.

Flying creatures - bugs

Downloads

- Bugs zip

Note: requires Framsticks v2.10.10

Description

These are among the simplest creatures available in Framsticks. Because of their simplicity and uniform behaviour, they are used in artificial ecosystem experiments.

These 'bugs' are creatures that consist of three branches, encoded in f1 as X(X,X). On the latter two endings, we attach sensors. On the remaining branch, we attach two custommade motors: forward and rotation.

The forward neuron translates the whole creature some distance (dependent on its activation) in the direction of this stick.
The rotation neuron rotates the whole creature around the center of the creature.

Braitenberg vehicles

Downloads

- Braitenberg zip

Note: requires Framsticks v2.10.10

Description

Valentino Braitenberg's booklet 'Vehicles' has received much attention in the robotics community. It continues to inspire many people by showing that complex behaviour does not require a complex brain, but emerges from the interactions between the vehicle and its environment. The more complex its environment, the more complex its behaviour.

We've created these Braitenberg vehicles in Framsticks to demonstrate these principles.

The creatures use food-smelling sensors, and have two wheel-motors.

Related literature

V. Braitenberg, Vehicles: experiments in synthetic psychology, MIT Press, 1986.