Fantom is an implementation of a generalized version of the popular board game Scotland Yard (or its Czech version Fantom staré Prahy). This implementation is designed especially for testing AI and machine learning algorithms - there is no graphical user interface, but it uses a client/server architecture with a simple and well-documented interface between the server and the clients.
From the viewpoint of AI research, this game has several interesting features - when controlling the detectives, the agent must successfully deal with uncertainty in the game, and it must carefully plan and coordinate its movement with respect to available resources. For Mr. X, the agent needs to avoid the detectives, without being too predictable to be found easily.
Fantom was implemented by Jiří Iša and me for the Seminar on Applied Artificial Intelligence at Charles University, Faculty of Mathematics and Physics, but the source code is available under the GPL license, and it can be downloaded from BitBucket.
