An Interview with Prof. Anne Lange of the University of Luxembourg
It is not every day that we have the opportunity to talk with passionate and talented researchers in a certain field, so it was a pleasure to interview Professor Anne Lange. Professor Lange is an associate professor in Logistics and Supply Chain Management at the University of Luxembourg.
Her research interests involve operation research (OR), transportation, network analysis, and supply chain finance. She is fascinated by real-life applications, often working in cooperation with companies, such as DB Schenker, DHL, and Cargolux. She received her doctoral degree from the University of Cologne, and was a postdoc at TU Eindhoven. Professor Lange works with AIMMS on a regular basis, mainly applying network analysis tools for educational purposes.
In this interview, we discuss a little about her research experiences, the recent trends in the OR and optimization field, and how AIMMS improved her lectures in network analysis.
Can you tell us about yourself, how you came to be involved in this field, and your research interests?
I have always been fascinated by organizing things, even before finishing school - I might be a good become a wedding planner after all. No, seriously, I have a master’s in information systems where I learned programming. I also got involved in two electives: operations research and logistics. During that time, things just developed naturally, one project after the other, and I never had a reason to leave.
At the moment, my research interests can be condensed into two broad streams: aviation and supply chains. I enjoy studying networks and how things interact. I do a lot of research in aviation because there is a lot of data, and you can run some analysis on the network and evaluate how the nodes interact. Although the general theme is transportation networks, for some reason aviation got me really hooked up.
I also work on supply chains as a form of network, which I think is a more strategic kind of problem. I’m interested in studying how firms work together. This encompasses subcontracting, outsourcing, how to work with your supplier, what is the effect of something new happening in the supply chain, etc. For instance, we just completed a large study about the introduction of 3D printing in supply chains, which is slowly becoming part of serial production. In this case, we were interested in how 3D printing can change the structure of the supply chain itself.
From looking at your publications, it looks like you have a pretty wide scope of interests in your field. How do you decide on a new research project to take on?
I think it is really about grasping opportunities, you know? The longer I work on this the more I know what I want to do. For instance, I started doing more and more aviation, which gave me more context about it, so I started having more ideas about the field. When you start talking about a topic with a partner, and you realize that the company is interested, you want to work with this company, the question is motivating, and you have colleagues that also want to join in, it simply becomes a good package. And the more you dig in a field, the more you extend what you already finished, applying it in different fields.
So, the important things are: having an interesting question that motivates me, having colleagues whom I can trust, and if possible, a real application. I mean, theoretical work is great, but it is great to have an applied basis.
The important things are: having an interesting question that motivates me, having colleagues whom I can trust, and if possible, a real application.
Sounds like it is a demand-based research approach. Do you have a specific decision-making process to accept or reject research propositions?
You have to make sure that certain demands appear. And, of course, you don’t say yes or no in a day. You discuss it with people for a bit, you build an initial understanding of the subject, and then you meet again, and then, at some point one needs to decide: if I realize there is no drive or motivation, or I see that I can’t contribute in any way, then there is no point in continuing. And sometimes there are too many opportunities at the same time, and I just have to select the best ones.
Would you say that one of the appeals of your field is that you are always working with different industries and applications?
Yes, and I love visiting and seeing what is happening in industry. For instance, today we got to see a lab where our colleagues are researching metal 3D printing, and I enjoyed going in there and asking “What is this?” and “What are you doing there?” And it is super neat when, during an interview with the companies, we explain what we want to do, and they ask “Do you want to see it?” Well, I’m not a mechanical engineer, I don’t know anything about the processes of 3D printing beyond what I can understand, so I answer “Of course, I want to see it!”, and then I get a tour of the lab to see the processes in first hand. Although it may have nothing to do with the interview, it is amazing and a great door opener.
Can you share some of the most inspiring projects you've been involved in during your career? Ones that were eye-opening in some way?
I think the best ones are always the ones you are currently doing. We recently started a really nice project with an airline, concerning their turnaround processes, and this is certainly a thing I like. And there is also the 3D printing study that we finished, which may actually lead to new projects moving forward.
But probably the most eye-opening project to me, particularly, was my own Ph.D. project. Maybe because it was my first - the first time I was “thrown into the water.” I had to understand how companies approach research questions, and that the research question by default is not something companies are interested in, so you need to frame it in a way that shows the contribution to both sides, nailing down where the business application is. Also, my Ph.D. research showed me that I wanted to continue working in this field.
Tell us a little more about your Ph.D. project.
It was on transportation, focused on less-than-truckload networks, in collaboration with a company. We were trying to understand what a good network for the European market would look like. Of course, you can calculate the optimum network in a certain scenario, but we were looking 25 years ahead, and that is when your optimization kind of stalls. You need to go into scenarios where it is not about optimizing the cost function, as we have no clue of how the input data is going to be 25 years from now in a certain part of Europe. So, we started talking about different network structures, and which ones were good solutions. We came up with general ideas of how these networks should be structured, and not exactly which plant location is optimal.
When did you first discover AIMMS and how do you use the software in the classroom?
Honestly, I don’t remember exactly when I first heard about it. I had the great opportunity to explore the classroom usage with a graduate from our Master program. We use it in a class on logistics management, where we teach network design.
Previously, we used Excel and Python for network design, but students spent a lot of time fighting with the technical details, whereas the class is not about coding. It is about understanding what drives a network design to be one way or the other, understanding the trade-offs, how adding a facility will impact the transportation costs, and what can you do when you have a set of constraints.
Working with AIMMS makes it a lot easier to get to that point. We went through some learning process on how to set up the questions, but it really goes down to giving the student the opportunity to play with a solution. Although it is easy to make a network and calculate the optimal solution, the pleasure lies in exploring further. For instance, what would you do if you had different transportation modes? What if your customers demand shorter lead time requirements?
It really goes down to giving the student the opportunity to play with a solution. Although it is easy to make a network and calculate the optimal solution, the pleasure lies in exploring further.
And how is the training process? How difficult is it for the students?
We offer some training sessions to get the students familiar with the tool. We also use some of the online material, so that the students can prepare for the sessions and learn much faster. First, we give a general introduction, and then they work on the tool for two or three weeks—we individually help them with questions and debug problems, always guiding them through the process. In terms of difficulty, I think that setting up the configurations is a bit hard at the beginning, as they are not computer scientists. But in the network design process itself, I think it is a very intuitive tool.
You'll teach a course on supply chain management at the upcoming AIMMS Campus event. Can you tell us a little about it?
What I want to do is an overall introduction to supply chain management, in particular to the quantitative elements, the way you make decisions based on data, and where some of it can be answered with AIMMS.
Basically, there are two things I want to make sure I go through, and one is the strong interdependencies between the different elements in the supply chain. For instance, forecasting has an impact on inventory management, which also impacts transportation; and once a decision is made, you may have to stick with it if things change. So basically, the trade-off that you have between the different elements of the supply chain. And the second is: how much data you actually need, how far changes in data are going to change your decisions, and how crucial it is to make sure you have good data.
Do you have any advice for students studying in the OR field?
From my perspective, I think it is extremely valuable to think about applications. Learning the tools that you need in order to solve relevant problems, I think, is the most important thing. Because sometimes our field goes in the direction where it is about the pleasure of solving a complex problem, which is a highly academic challenge. But we should really invest in contributing value to industry and society in general. Rather invest your time to apply existing knowledge to solve challenging problems and continue to build new knowledge on it.
Sometimes our field goes in the direction where it is about the pleasure of solving a complex problem. But we should really invest in contributing value to industry and society in general.
How do see the future of Optimization and OR, both in studies and industry?
Sometimes I have the impression that, academically, we have achieved a lot already, in the sense that many highly relevant questions have been asked and answered, and we are just making small improvements, which isn't the way I would like the field to evolve.
On the industry side, computing power and availability of data sometimes overshadows optimization. I feel that companies have the idea that now we can predict everything, we don’t need to optimize anymore. I hope this converges at some point, and maybe this hype of just collecting a lot of data is going to dry out eventually and we go back to what type of data we actually need to improve to make better decisions. Optimization needs data for sure, but not every piece of data that is being collected is valuable.