Considering the German transportation growth prognosis (Verkehrsprognose fur die Bundesverkehrswegeplanung) for the year 2015 the rail traffic performance has to improve drastically in order to avoid possible traffic chaos on German roads. According to this it is especially necessary to encourage the competitiveness of rail traffic. Therefore the advantages of road traffic flexibility and punctuality have to be established comparably in rail traffic. Presently the railway management is based on a centralized supervision and dispatching of railway operation, which has to be optimized in order to handle the upcoming growth of transportation. This task can be performed by automated train path routing procedures based on mathematical optimizations. This paper shows a successful promising approach using genetic algorithms. The application of this stochastic optimization procedure requires a highly efficient simulation model of railway operations, which in the present case has been implemented by using Petri-nets. On the bases of a practical case study the possible fields of application for this prototypical implemented procedure of train path routing are presented. 1 Motivation The continuous economical growth in European countries during the last twenty years is closely connected with a nearly exponential increase of traffic volumes on road and rail. The forecast for the next 15 years is an expected near 50% increase of road traffic volume in Germany which, despite an improvement in road infrastructure, could lead to a collapse. One way to deal with this threatening forecast is to increase the transport capacity of rail transport. To achieve this aim in challengComputers in Railways IX, J. Allan, C. A. Brebbia, R. J. Hill, G. Sciutto & S. Sone (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-715-9 ing free market conditions, the punctuality and flexibility of transport by rail must be comparable to that by road. These goals are only achievable with very fast train traffic planning (flexibility) and optimal online dispatching (punctuality). At present, the main part of planning tasks is carried out by humans. With an increase in transport volume and decrease in available planning time, the complexity of decision making increases exponentially and is therefore no longer so manageable by human planners. Computer based support of this task is essential for rail transport companies. To deal with this task many mathematical approaches have been investigated during the last three decades [1, 2, 3, 4, 5]. Their common drawbacks are either too low precision or too high computational efforts. This is based on the fact, that the planning task is NP-complete and there is no deterministic method, which can find the solution with polynomial complexity. 2 Approach The actual planning process carried out by railway company management can be modelled as a control task consisting of two iterative control loops (fig. 1). Timetabling Dispatching Traffic process Request Time table Figure 1: Control loops in transportation process. In the frame of the first control loop the trains are scheduled according to the requirements of transport providers. This is an iterative process, with the aim of finding out the best compromise between available slots and customer requirements. Several factors must be considered: besides source-destination relationships, additional factors like prices, priorities, required punctuality etc are very important. All these factors have to be implemented on train paths and time instants in the form of a time table. This provides an input for the second control loop dealing with operative dispatching. The second control loop is responsible for online train control reacting on actual disturbances in the transport process. This often requires re-scheduling of the original time table, in order to minimise the consequences of operational conflicts. The main difference is that the first control loop represents a much shorter reaction time. While train planning can be performed within several hours or days, dispatching has to be effected within minutes or even seconds. Computers in Railways IX, J. Allan, C. A. Brebbia, R. J. Hill, G. Sciutto & S. Sone (Editors) © 2004 WIT Press, www.witpress.com, ISBN 1-85312-715-9 776 Computers in Railways IX