XRaptor is an environment for simulation of scenarios in continuous virtual multi-agent worlds. It is written in C++ for UNIX platforms with the X Window System and Motif 1.2. XRaptor allows studying the behavior of agents in different 2- or 3-dimensional continuous worlds. In the current version an agent is either a point in its world, or occupies a circular area or a spherical volume. It contains a sensor unit through which it obtains information about the outside world as well as an actor unit through which it performs actions. The agents are controlled by a user-defined control kernel. The control kernel determines an agent's action based on the sensory input. A typical control kernel is designed to maximize the agent's survival time. The user can construct an agent by implementing concrete control kernels derived from abstract classes (i.e. classes with pure virtual methods) provided by the XRaptor environment, whose pure virtual methods have to be overridden by the user.

Browse the manual to get an idea what XRaptor is all about and to learn how to install XRaptor.
Download the manual (pdf, ~718K)
Download the manual (ps, ~337K)
Download the source code (gzipped tar, ~284K)

If you are interested in multi-agent simulation and Artificial Intelligence, you might also be interested in information about

• RoboCup, the robot soccer world championship
• The Mainz Rolling Brains, our RoboCup soccer agent team