My research visit to Aachen

15th May 2015

One of the nice things about being in Europe is the accessibility to many different people working in my field. While there is a strong operations research presence in Australia, the population is not large enough to have the same density of researchers. I am definitely enjoying the travelling opportunities that are available to me in Europe.

I have been very fortunate and received an invitation from Marco Lübbecke to visit his research group in Aachen. The idea to travel to Aachen originated at the end of last year. The plan was to work on a column-and-row generation implementation and discuss general ideas about column generation research topics. From an initial discussion with Marco in Berlin, it felt like both sides would benefit from this visit. Well, that is what I was hoping for.

Work travel and strikes

I don't know what it is, but I seem to be finding the periods of time when transportation workers are going on strike. My travel to Bordeaux occurred during one of the longest Air France pilot strikes and when I attended the DIMACS Workshop the Lufthansa pilots went on strike. This time, the train drivers of Deutsche Bahn went on strike for a week starting the day I planned to travel from Berlin to Aachen. I don't travel for work very often, but in the last year there seems to be an over abundance of strikes occurring when I want to use a particular service.

This strike was not much of a problem. I was expecting the worst. My trip to Bordeaux (more accurately the trip home) was much more affected by the pilot strike than this trip was affected by the train drivers strike. I was concerned, so I contacted Marco to ask whether it would be better for me to postpone my trip until after the strike. His response was very relaxed. The suggestion was to just travel on a train – any train – departing after my original booking. Because of the strike, the ticket inspectors are much more lenient and will not care than I have a specific train ticket. It was supposed to be reassuring, and it was. I just did not have the experience to know that things were going to work out that well.

My travel from Berlin to Aachen went ahead without much trouble at all. Arriving at the Berlin Hauptbahnhof I was surprised by how fewer people there were. I was expecting large numbers of people waiting for later services after their particular train has been cancelled. Then getting on the train, again it was surprising that there were spare seats available. I was able to travel to Aachen without evening knowing the train drivers were on strike. The only thing that changed was my departure from Berlin was 45 minutes later than originally planned. Not much of a problem at all.

Getting into work

I arrived at the university on Tuesday afternoon. Immediately upon my arrival I met with Jonas Witt and Sarah Kirchener. This immediate dive into work really framed how we were going to spend the next two weeks. We started talking about column-and-row generation and the implementation Sarah has been working on. I have employed column-and-row generation before, but there are many different generic approaches that have been developed. Specifically, the approach by Sadykov and Vanderbeck is very different to the method I employed for the integrated airline recovery problem. It is the former approach that was implemented by Sarah for a job scheduling problem.

This was a great experience for me. First, I had not implemented the Sadykov and Vanderbeck column-and-row generation approach – even though it motivated my research in this area. Second, a focus of this week was to look at cutting planes for column generation and column-and-row generation, something that I don't have a strong knowledge of. Since cutting planes for branch-and-price is a research topic of Jonas, there was much that I would be able to learn. So it was looking to be a great two weeks of learning and research discussions.

Giving a seminar

There was no slowing down. I arrived on Tuesday afternoon, by Tuesday evening my seminar was scheduled for 11am Wednesday. It was a good thing that I wanted to be prepared for this work trip. Otherwise it would have been a very stressful Tuesday evening. The only thing that I needed to do in preparation of my seminar was write the date on the first slide as "6th May".

The seminar I delivered was on an application of column-and-row generation to the integrated airline recovery problem. I felt that this was appropriate given that one motivation for travelling to Aachen was to discuss column-and-row generation. I felt very comfortable with the presentation. I have delivered versions of it at other times over the last two years. So the work felt very familiar. Unfortunately I fell into a common Australian problem of talking too fast. Given that my audience were not native English speakers, I feel that I may have lost some along the way. This is definitely something that I have to be more mindful about in the future with my delivery.

The questions that I received at the end of the presentation were really interesting and motivating. The audience appeared to have some interest with this work and were able to ask very relevant and considered questions. It rekindled an interest in this type of research. It will definitely go on my projects list. Hopefully reach the top soon and receive an appropriate amount of attention.

Great social environment

Something that I have found in the different research groups I have visited is a good social culture within the office. While completing my PhD in Australia, there was a feeling in the department that while you are at work, you work, and play time is for when you finish. It was still common to have social interactions with others in the department, but this generally occurred after work and at external social activities. Sitting down for a coffee or lunch break as a group was not as common. In Germany, it appears that going for lunch is a sacred tradition and it is something that you should try to be a part of. Admittedly, I don't partake very often at ZIB, but I do enjoy the culture around it.

The research group in Aachen was not overwhelmingly large. Hence, everyone in the group knew each other well. Even across different research groups in the same department. I discovered this after being invited to a house warming party. It was a great opportunity to meet and talk socially with the members of the research group I was visiting. You get such a different perspective in a social environment compared to meeting just in the office. As usual, this was a very positive experience.

The outcomes

I really wanted to go into this with very clear goals in mind. I even spent a bit of time trying to determine what is a worthwhile and achievable set of goals. As with most things in research, it is always necessary to keep your goals flexible and be prepared to adjust them as time goes on.

During the week a few things happened. First, it was not immediately clear how to achieve some of the goals, while others were much easier to address. In this respect, I spent a bit of time implementing some work with Sarah, which hopefully makes a difference to the algorithm performance. Second, the time spent on some tasks took a little longer than originally planned. Third, people are busy and you also don't want to spend all of your available hours discussing research. I found that the time required to complete all tasks was not really there. However, in regards to this latter point, one of the objectives of this visit was to commence a collaborative project. There was much progress towards this goal.

Something I find important is to reflect on outcomes. For this research visit, I feel that this was very beneficial for me to meet some interesting and talented people. Further, the work discussed was very interesting and it felt exciting to be engaged with it. Professionally, I believe that the discussion will lead to some promising research. Hopefully this will be something that can be published in the future. Apart from the professional outcomes, this was a very enjoyable experience and once again motivated me to delve into my research at ZIB.

The wrap-up

Aachen was a very nice and fun city. I enjoyed the closeness and comfort that I felt while exploring the city. I had the impression of an overall sense of community. This, I believe, comes from size of the city and a concentration of people and services. In particular, I enjoyed the student presence in the city and the energy that comes with that.

The work was a lot of fun. Many times I felt lost in our discussions, other times I felt really involved. This is all part of learning. The opportunity afforded to me for such work trips is not lost. I am grateful to the people who provide these opportunities and it is something that I endeavour to make the most of.