I obtained my PhD from Eindhoven University of Technology for research on Single Machine Bicriteria Scheduling problems; I conducted this research at the `Centrum voor Wiskunde en Informatica' (Centre for Mathematics and Computer Science) in Amsterdam under the guidance of Jan Karel Lenstra. After that I continued this research on machine scheduling problems at Eindhoven University of Technology (postdoc), Princeton University (postdoc) and Eindhoven University of Technology again. In Eindhoven I obtained several breakthrough results on for example devising algorithms with exact competitive ratios for on-line algorithms and application of column generation algorithms for scheduling problems.
After a number of years I moved to the computer science department at Utrecht University, where I made the switch of moving to applied research. Together with Marjan van den Akker and several Msc and PhD students I study real life planning problems that are actually used. In this way I initiated a research line on finding robust solutions, which remain feasible in case of small disruptions. Together with Marjan van den Akker I started a small Operations Research group, which forms the core of the recently established AI and Mobility Lab. We have conducted a large number of relevant, useful research projects together with major companies. Examples of successful projects are
- Gate assignment at Amsterdam Airport Schiphol. Here the goal is to assign planes to gates subject to size and security constraints, such that the robustness is maximized. This is part of the PhD-project of Guido Diepen.
- Planning bus schedules and driver schedules for Qbuzz, which is a major bus company that serves public transport in a large part of The Netherlands (for example in Utrecht under the name U-OV). This is part of the PhD-projects of Marcel van Kooten Niekerk and Jan Posthoorn (ongoing).
- Planning service sites for NS (Dutch Railways). The goal here is to construct feasible shunting plans for shunting yards at which also all kinds of services, like cleaning and technical inspections, take place. This is part of the PhD-project of Roel van den Broek; we have devised the first algorithm that is capable of solving real-life instances.
The success of the project for NS has led to a number of new projects within the kickstart AI initiative, which has been established by four big companies (NS, KLM, ING, Ahold-Delhaize) to boost AI research in The Netherlands. We have been selected to
- conduct research for improving the planning of city hubs, which consist of a number service sites. This project consists of two PhD-positions and a secondment of one day per week at NS for Han Hoogeveen. It has started in September, 2021.
- conduct research for improving the planning of KLM. This project also consists of two PhD-positions and a secondment of one day per week at KLM for Marjan van den Akker.
Project 1 with NS has very recently be extended as part of the LTP ROBUST plan of ICAI, in which we collaborate with NS, ProRail, and TU Delft in the so-called Rail Lab. For the 乐鱼后台 this has led to an additional two PhD-projects. Finally, we have just obtained a PhD-project together with Egon van den Broek (乐鱼后台) on making rosters for NS personnel in which we take human characteristics into account as well.
As can be concluded from the success mentioned above, I think that it has been a very good move to start working on applied research in which real-life problems are researched and solved. After a slow adoption this research has no received full appreciation. The key to success is found in the following slogans:
`Do more with less' and `Why conduct research if you are the only one who is interested in the outcome'. To make the distinction clear between our work and the work done by consultants: `We are not interested in the color of the hood, but in what is running beneath it'.