The design of new railway infrastructure is a complex planning process in most countries today due to a multitude of requirements. From an operational point of view new infrastructure basically has to fulfill the needs defined by customers. To this end passenger traffic is often organized in an integrated timetable with well defined arrival and departure times at major hub stations. So far there is no automated tool available to help in determining a minimum cost infrastructure fulfilling all the requirements defined by a timetable and the operation of the railway system. Instead, this task is typically carried out manually, based on graphical design, human experience, and also intuition. In our work we model this planning task as a combinatorial network optimization problem, capturing the most essential aspects. We then present a constructive heuristic algorithm that makes use of a dynamic programming procedure for realizing individual commercial stops. Computational experiments on instances derived from real scenarios indicate that the suggested approach is promising and the analysis of obtained results gives useful hints for future work in this area.

railway infrastructure design; integrated timetables; combinatorial optimization; dynamic programming; heuristics