Crogai is a mix of crowd simulation, artificial intelligence, and genetic algorithms. It is possible to set up simulations and test different AI in a full 3D environment with a small physics engine, terrain, collision handling, etc. Crogai is inspired by the OpenSteer projet and uses steering behaviors as building blocks for the AI. The agents can graze, hunt each other, and emit spores for reproduction. A genetic algorithm allows the agents to evolve, and let them decide whether to do so or not. In addition, the examples provided by the project make nice screensavers.
The Kernel-Machine Library is a C++ library to promote the use and progress of kernel machines. It is both for academic use and for developing real world applications. The Kernel-Machine Library draws heavily from features of modern C++ such as template meta-programming to achieve high performance while at the same time offering a comfortable interface. It enables compile-time selection of specialized algorithms on the basis of data types: for example, the specific case of a SVM in combination with a linear kernel can be computed by a specialized efficient algorithm.
CharGer is a conceptual graph editor intended to support research projects and education. It currently is primarily an editor to create visual displays of graphs. It is deliberately and explicitly a research tool meant for conceptual graph researchers to explore implementation issues in conceptual graph interfaces. Using the software will require some familiarity with conceptual graphs, including knowing about concepts and relations, type hierarchies, and type/referent pairs. Knowing about actors will also be very helpful.
SeSAm provides a generic environment for modelling and experimenting with agent-based simulation. As there are freely configurable instruments for gathering data and scripting options for constructing simulation experiments, SeSAm is a highly valuable tool for MAS simulations, especially for complex models with flexible agent behaviour and interactions.
Enki is a fast 2D physics-based robot simulator. It is able to simulate cinematics, collisions, sensors, and cameras of robots on a flat surface. It also provides limited support for friction. It is able to simulate groups of robots hundreds of times faster than real-time on a modern desktop computer.
Teem is a software framework for conducting artificial evolution experiments. It particular, it provides lots of facilities for evolutionary robotics. It can be combined with a variety of libraries and programs to form a powerful platform for anything from evolution of simple Braitenberg behaviour to complex steady state evolutionary experiments involving groups of robots.
The RL-Glue is intended to provide a foundation for building benchmarks for reinforcement learning (RL) agents. A secondary goal is to support RL competitions in which agents are compared in their performance on new problems that are only revealed at the time of a competition. The current version of the interface software supports direct function call communication between agents and environments written in C. The interface software also provides a standard mechanism for communication between agents and environments written in different programming languages.