50 years of OR in railway timetabling and rolling stock planning

Abstract

This survey paper discusses the literature on Operations Research (OR) models and algorithms for railway planning in the last decades. Since infrastructure and rolling stock are two resources that are both very capital-intensive and characterize the railway system, we focus on timetabling and rolling stock planning problems. For timetabling, we also classify the literature among two dimensions, namely the decision level (strategic, tactical, operational) and the type of network infrastructure (dependent routes, independent routes). We also discuss robustness aspects in both planning problems. We focus the discussion of the literature on the applicability of the models in the European context, where different types of trains (high-speed passenger trains, high-frequent suburban trains and freight trains) often share the same tracks and the organization of the railways is usually split between an infrastructure manager and one or more railway undertakings operating the trains. We conclude the paper with some challenges and future research directions.