ERIM Home

In her dissertation Sha Zhu presented three contributions to the problem of obsolescence of large spare parts inventories for capital goods. First, a new method based on extreme-value theory was developed to aid companies in forecasting the spare parts demand distribution. Next, the inventory control problem for on-condition maintenance and shutdown maintenance were analyzed. A new approach for joint forecasting and inventory control based on probabilistic information on the maintenance plan is proposed. It was found that the value of this plan is significant in preparing the repair shop by catching the irregularity and lumpiness of spare parts demand. Finally, a model is established to model the spare parts ordering problem against the background of shutdown maintenance project planning. Decision makers need strategies which consider the interdependence of maintenance activities. This new stochastic programming approach is able to give much better advice than traditional methods and hence meets the requirement of real-life shutdown projects.

In his dissertation 'Public Transport and Passengers: Optimization Models that Consider Travel Demand' Johann Hartleb presented integrated models to optimize public transport services while estimating the corresponding passenger choices. The first study compares different timetable evaluation functions for consistency and gives further motivation for the integration of passenger choice models into optimization models. In the next two studies, he presented novel optimization models with integrated demand estimation for the steps of timetabling and line planning, respectively. The resulting public transport services are designed for the passenger demand they generate. The second part of this thesis deals with new and more flexible forms of public transport: mobility on demand. In order to assess the consequences of large-scale on-demand services on cities and regions, travel demand models need to be extended to determine the service level of on-demand services. Both studies in this part present solution algorithms for a vehicle scheduling problem of on-demand services to estimate the required vehicle fleet size and distance traveled