Disruption Management

Latest Developments – Evelien van der Hurk (CV)

Currently, Evelien is in the final stages of finishing her PhD based on the ComPuTr project. She works on the themes of Passengers, Information, and Robustness. Central is the question how the increased information on passenger behavior (such as resultant from smart card data), and the newly available opportunities to communicate with passengers, can be used to improve service during major disruptions. The resulting research can be divided into two parts: the first studies how operators can obtain detailed insight into the passenger flows and demand based on the newly available data. The second investigates how to improve disruption management techniques using the available data, and include the option for direct communication with passengers. Key aspect is that realistic assumptions on passenger behavior are included, and sensitivity to different passenger behavior assumptions was studied. This lead to new optimization based models to minimize passenger behavior, which showed to be (very) successful for realistic case studies based on data of both Netherlands Railways and the Massachusetts Bay Transportation Authority, USA.

As part of her PhD, Evelien visited Nigel Wilson and Haris Koutsopoulos at the Transit lab at the Massachusetts Institute of Technology, in the fall of 2013.

The following projects resulted from her PhD:

Project 1: Passenger Route deduction (In IEEE Transactions on ITS)
Deducing passengers’ route choice from smart card data provides public transport operators the opportunity to evaluate passenger service. Especially in case of disruptions when route choice models may not be valid this is an advantage. This paper proposes a method for deducing the chosen route of passengers based on smart card data and validates this method on a real life data set. The method reaches an accuracy of about 90 percent, also in case of disruptions. Moreover, it is shown
how this method can be used to evaluate passenger service by a case study based on a real life data set of Netherlands Railways, the largest passenger railway operator of the Netherlands.

Project 2: Forecasting passenger behavior in Disruption Management (In Rail Copenhagen Conference Proceedings )
Automatic Ticketing Systems like smart card data provide detailed data on passenger behavior that was not available in this detail before. Such information could assist operators in obtaining insight in the demand for travel. In case of disruptions, information on where currently passengers are, and where they are traveling to, is essential to provide good service, for instance during (planned) disruptions (project 3 and 4). A two-step methodology consisting of Network Reduction and forecasting leads to time and origin destination specific forecasts of passenger journeys. Immediate insight into how passengers are affected by a disruption results from the Network Reduction. Real-time forecasts of passenger behavior are derived based on historic information from smart card data. Results are based on a 10 months’ real-life smart card data set of the Dutch passenger railway operator Netherlands Railways (NS).

Project 3: Shuttle Planning for Line Closures (ERIM technical report, submitted to Transportation Science )
Urban Public Transport systems must periodically close certain links for maintenance, which can have significant effects on the service provided to passengers. In practice, the effects of closures are mitigated by replacing the link with a simple shuttle service. However, alternative shuttle services could reduce inconvenience at lower operating cost. This paper proposes a model to select shuttle lines and frequencies under budget constraints. A new formulation is proposed that allows a minimal frequency restriction on any line that is operated, and minimizes passenger inconvenience cost, including transfers and frequency-dependent waiting time. This model is applied to a shuttle design problem based on a real world case study of the MBTA network of Boston (USA). The results show that additional shuttle routes can reduce passenger delay in comparison to the standard industry practice, while also distributing delay more equally over passengers, at the same operating budget. The results are robust under different assumptions about passenger route choice behavior. Computational experiments show that the proposed formulation, coupled with a preprocessing step, can be solved faster than prior formulations.

Project 4: Disruption Management with Robust Passenger Guidance and Rolling Stock Rescheduling
Public transport networks like that of Netherlands Railways cope with several major disruptions a week, e.g. due to malfunctioning rolling stock, infrastructure, or accidents. The duration of the disruption is uncertain at the start. Still an operator needs to adjust his logistic planning, and passengers need to reroute, unknowingly of the length of the disruption. Passengers’ best route depends on both the available capacity, as well as the duration of the disruption. The available capacity depends on both the rolling stock schedule, defining the available number of seats per train trip, as well as the route choice of other passengers. The operator is the only one that has insight into both. By providing route advice to passengers the Operator can help them avoid sections that are overcrowded and well cause more delay. By adjusting the rolling stock schedule the operator can aim to provide sufficient seats for the demand, to the extend this is possible.

A novel optimization based approach is presented that proposes both advise to passengers and a new rolling stock schedule minimizing anticipated delay under realistic assumptions on passenger behavior, and given an uncertain length of the disruption. Thus is finds robust solutions for both given a set of possible disruption lengths.

Results show that the provided information and route advice to passengers is more important than the initial rolling stock schedule, supporting the opinion of the public, and the statement of NS, that the passenger should come first. Current results are promising that our model will (in theory) be able to support NS in this respect in the future.

 

General Project Description: Disruption Management

The ComPuTr projects consists of two tracks: Disruption Management and Revenue Management. This page is about the first track, the PhD project of Evelien van der Hurk. Research Proposal

Recent technological advancements leading to the introduction of smart cards and the wide adoption of smart phones present ample opportunity in public transport for understanding passenger behavior in far more detail than was previously possible. These technologies however do not only provide detailed data, they also provide an immediate link between Public Transport Operators (PTOs) and passengers, enabling instant communication and distribution of information thanks to the omnipresence of wifi.

Smart Card Data

About smart card data: In the Netherlands, smart card data contains the start and end point of every journey. For train journeys, it contains the time of check in (ci), the station of ci, the time of check out (co) and the station of co. Because this data is at the station level, multiple routes may be available to get from start to end station within the ci,co times. Conductor checks may assist in linking ci,co to a specific route.

This project is about how both passengers and PTOs can profit from these new technologies. It focuses on disruption management, in specific on the new opportunities for increasing system performance by actively including passengers enabled through increased information on passenger behavior and the possibility of instant communication.

The main research question is: How can a change in passengers’ travel strategies enabled through Informedness contribute to better disruption management?

This question is linked to three themes: Passengers, Information and Robustness. The first concerns analysis and prediction of passenger flows. The second is about modeling passenger behavior and the influence of information. Finally, the theme Robustness focusses on the sensitivity to uncertainty of the quantitative model resulting from the themes Passengers and Information.

We combine theory from Complexity Theory with Operations Research techniques in the study of these themes. The focus in this project is on quantitative modeling for the solution of this problem.

Research is conducted in close collaboration with Netherlands Railways.