Picture of Twan Dollevoet Welcome to the homepage of Twan Dollevoet

Latest News

Personal Data

Publications

Research


News highlights

More news can be found in the archive.

Personal Data

Twan Dollevoet
Econometric Institute
Erasmus School of Economics
Erasmus University Rotterdam
PO Box 1738, 3000 DR Rotterdam, the Netherlands
Office: Campus Woudestein, E-building, T-floor, usually in ET-36
Tel.: +31 (0)10 754 5277
Email: dollevoet@ese.eur.nl

See also my LinkedIn profile, my EUR website, my ESE website, my ERIM website, my ORCID website, and my Google Scholar profile.

Publications

Accepted and published journal papers

Refereed conference contributions

Edited volumes

Chapters in books

Ph.D. thesis

Professional contributions

Working papers


Research

I am an associate professor at the Econometric Institute of the Erasmus School of Economics, Erasmus University Rotterdam. My research focuses on applications of combinatorial optimization, most notably in transportation. For my Ph.D. thesis, I have developed several extensions of the classical delay management model. Directly thereafter, I have been working on online delay management and integrated disruption management within the European project ON-TIME. My current research interests include public transport optimization, vehicle routing, and kidney allocation. Some of my projects are described below.

Current projects

Together with Roby Cremers, Michal Mankowski (NYU Grossman School of Medicine), Joris van de Klundert (Universidad Adolfo Ibanez), and Albert Wagelmans, I am working on kidney transplantation policies. In particular, we use Monte Carlo simulation and advanced optimization methods to assess the performance of various kidney allocation policies.

Together with Danny Zhu and Dennis Huisman, I am working on rolling stock planning. In particular, we are integrating various stages in rolling stock rescheduling. As a first step, we consider the routing and scheduling of moving rolling stock units that are coupled to or uncoupled from rolling stock compositions from and to the shunting yard.

Together with Bart van Rossum and Dennis Huisman, I am working on railway crew planning. Currently, we are working on a new approach to crew scheduling, incorporating fairness at an individual level and postponing the crew scheduling decisions to a later stage. In such an approach, so-called template roster are generated well in advance, while precise duties are determined closer to the day of operation.

Together with Pedro Munari and Remy Spliet, I work on a generalized formulation for Vehicle Routing Problems (VRPs). The generalized formulation includes the arc-flow and set partitioning formulations as special cases. A report on this research can be found among my working papers. Together with Diego Pecin and Remy Spliet, I am studying all kinds of extensions of this idea.

Past projects

Together with Rowan Hoogervorst and Dennis Huisman, I have worked on rolling stock rescheduling. First, we have incorporated delays and added rescheduling flexibility into existing approaches. Second, we have developed a heuristic for rolling stock rescheduling. Finally, we have integrated the scheduling of shunting drivers into rolling stock rescheduling. Rowan defended his Ph.D. thesis on October 14, 2021. His thesis can be downloaded here. Rowan is now a post-doc at DTU, Copenhagen.

Together with Thomas Breugem and Dennis Huisman, I have worked on railway crew planning. In particular, we have incorporated fairness aspects in the traditional crew rostering approaches. Thomas has defended his Ph.D. thesis on January 24, 2020. His thesis can be downloaded here. After obtaining his PhD, Thomas first moved to INSEAD, France. Currently, he is an Associate Professor at Tilburg University.

Shyam Sundar (DTU Denmark) has visitied Erasmus University Rotterdam during the Spring and Summer of 2019. Together with Shyam, Dennis Huisman, Richard Lusby, Jesper Larsen, and Morten Riis, we have developed an ALNS algorithm for the problem of simultaneously scheduling electrical buses and drivers. Integrating the electrical vehicle and crew scheduling leads to significant improvements with respect to a sequential approach. Our work is published in Computers & Operations Research and can be found here.

From November 2016 to November 2017, dr. Chongshuang Chen from Southwest Jiaotong University has visited Erasmus University Rotterdam. Together, we worked on a novel algorithm for the train formation problem. We developed both an exact algorithm and a tree-based decomposition heuristic. A paper on our research is published in Transportation Research Path B: Methodological and can be downloaded here

Lukas Bach, a Ph.D. student at Aarhus University, has visited Erasmus University Rotterdam several times since April 2013. Together with Lukas and Dennis Huisman, I have worked on the integration of train timetabling, engine routing and crew scheduling for a real-world case of DB Schenker, Rail Scandinavia. By allowing for some flexibility in the timetale, the overall costs of the timetable and crew schedule could be reduced by 6-8 percent. A paper on this research has been published in Transportation Science and can be found here.

As a Ph.D. student, I have collaborated with some of my fellow Ph.D. students. Together with Willem van Jaarsveld, I have looked into an algorithm to solve multi-item inventory problems. We have written a paper on this project that has been published in Omega. Together with Lanah Evers, I have been working on robust solutions to the orienteering problem. Our paper on this research is published in a special issue of Annals of Operations Research. Together with Kristiaan Glorie and Theresia van Essen, I have studied the Tray Optimization Problem (TOP). Here, sterile intruments must be grouped into nets and are then assigned to surgeries. A paper on this research is published in European Journal of Operational Research.

When I was a Ph.D. student, I have worked part-time on crew scheduling at Netherlands Railways (NS). In cooperation with SISCOG, we have developed an algorithm that can tackle the crew scheduling problem for a whole week at once. One of the main results is a crew scheduling algorithm that can solve complete week instances both for drivers and for guards. A paper on this algorithm has been published in Public Transport.