Research project

General Description
The problem of finding optimal time tables comes in various flavours: schol time tabling, rostering, production planning or course scheduling at universities. Because of the complexity of the problem automated generation of solution is required.
In the last years several approaches have been sugessted, among them genetic algorithms, tabu search, constraint programming or heuristic methods. But the implementations and data structures of these approaches are different making it hard to try them out or even compare their results.
Starting in the year 2002 a uniform framework to develop, test and evaluate time tabling algorithms and problems has designed and implemneted. The formal description of problem and algorithm is written in a specification language based on the XML-standard. This description can be interpreted and executed by a framework written in Java. For a given problem a valid solution can be generated using one of the implemented algorithms.
Our next step will be the implementation of serveral algorithms the evaluate their ability to solve certain problems and their performance shown when applied to these problems.

We also intensified our school time tabling efforts by adding additional constraints to our database, especially for german primary and secondary schools.