Anita Schöbel: Integrated optimization of public transport

How to plan a public transportation system such that it is well-designed for the passengers and at the same time not too costly?

In theory and in practice, planning of a public transportation system is usually done in a sequential way: After the network design, the bus or train lines are established. Based on the line plan, a timetable is set up, and finally the vehicles’ schedules and the drivers’ schedules can be planned. From an optimization point of view such a sequential planning procedure can be regarded as a Greedy approach: in each planning stage one aims at the best one can do. This usually leads only to suboptimal solutions for the whole system. On the other hand, many of these single steps are already NP hard such that solving the integrated problem to optimality seems to be out of scope.

In this talk we argue that public transportation benefits from an integrated planning, and we sketch exact and heuristic solution approaches on how to tackle sequential processes in an integrated way.
Exact solution approaches include an analysis of the problem in star-shaped graphs and first ideas on decomposition approaches. For designing heuristics, we propose an eigenmodel for integrated

Evelien van der Hurk: Passengers, Information & Disruptions

Major disruptions cause the current logistical schedule of the operator to be infeasible. Adjusting this schedule to the disruption is a complicated planning problem. Passengers will adjust their journeys to the new schedule, and may need to adjust their route choice due to the route choice of other passengers in case of capacity shortages. Therefore the passenger service results from a complex interaction between passengers themselves, and between passengers and the schedule.

This presentation will discuss a new models for improving passenger service in case of major disruptions by adjusting the rolling stock schedule while anticipating passenger’s reactions, and also by supporting passengers during disruptions through the provision of route advice. If time allows, a link will be made with the a study on passenger behavior based on new data sources available generated byautomated fare collection systems such as the Dutch smart card.

Gert-Jaap Polinder: Robust Periodic Timetabling

In railway operations, there are disturbances every day. It is almost certain that disturbances will happen, but we do not know where and when. In the design of a timetable, you want to take this into account. A method to do so is shown in this talk.

In this talk, we will motivate why we want to find a robust timetable. We will show where this method is positioned with respect to other literature on robust timetabling. We will show the method we used to come up with a robust timetabling and provide some computational results.

Kai Nagel: Agent-Based Simulation for Transport Analysis

This presentation will explore the triangle between public transit, car-based mobility, and emissions/carbon footprint in the context of the emerging technology of autonomous and electric vechicles.  The method for this will be agent-based simulation, where each person and each vehicle is represented by a synthetic avatar, whose movements will be traced throughout the day by the simulation.  Several policy-relevant observations emerge, such as that in urban areas a replacement of all individual cars by a fleet of autonomous and electric vehicles would be both economically and technically feasible, and that carbon footprint reduction needs to concentrate not so much on urban traffic but on long-distance commuting and on freight.