Interview: Michel Mandjes, professor of Applied Probability
1 November 2022
Networks such as traffic, communication and energy networks form the backbone of our society. These networks are becoming increasingly complex.
‘As a result of this increasing complexity, there is a greater chance over time that something will go wrong or that a network will become unstable’, explains Mandjes. ‘As scientists, but also across society as a whole, we must have the courage to think creatively in order to ensure that these networks, which are increasingly subject to intensive use, will continue to be used efficiently in the future while still remaining reliable. That's why the research we are conducting within NETWORKS is so important.’
NETWORKS is a 10-year research programme, funded with a Gravitation grant of 22.7 million euros from the Netherlands Organisation for Scientific Research (NWO). In addition to the UvA, Eindhoven University of Technology (TU/e), Leiden University (UL) and the Center for Mathematics and Computer Science (CWI) are involved.
Mandjes: ‘We wrote the research proposal with the common thread that we really wanted to tackle this topic together, and that helped us to get this project off the ground. And these are not just empty words; all the parties are genuinely willing to learn about other disciplines. For example, we are increasing the interaction between principal investigators. This sometimes works well already, to be honest, but it is often a challenge because senior researchers all have their own agenda.’
Scientists with their own agenda? What’s new, you might say? As a project leader, this means that you must be able to be flexible. Mandjes has been able to find his own way here: ‘I try to be an invisible leader; my aim is to listen to what people want themselves, certainly without forcing them to go in a certain direction. Within these margins, it still seems possible to do all kinds of things to achieve a great deal of synergy.’
This approach seems to be working well. The research within the broad NETWORKS programme focuses on the stochastic (the outcome depends on chance or coincidence) and algorithmic (instructions are used to achieve a goal) aspects of networks. The focus lies on modelling, understanding, controlling and optimising networks that are complex and highly volatile. The programme comprises two branches: ‘Network structure’, which is more about describing networks, and ‘Network shaping’, which is more about the design of the networks.
One example of deep and challenging research takes place within ‘Network structure’ when scientists look at the limits of graphs. A graph is a schematic network of points, nodes, paths etc. You can allow such a network to grow, subject to certain regularity requirements, after which you can investigate the graph limits. In recent work, Mandjes has been examining the stochastic behaviour of graph limits, building on the work of people such as mathematician S.R.S. Varadhan, winner of the Abel Prize or the unofficial Nobel Prize in mathematics.
Although this is the fundamental aspect of the programme, the NETWORKS scientists have increasingly started to look at concrete applications. ‘We seek to connect with fields such as social sciences and economics, to ensure that our knowledge is well received’, says Mandjes. ‘These types of collaborations are not self-evident because every field of study has its own language. First you look at a particular topic from your own experience, then you have to talk a lot and adapt your models in order to achieve a result that is relevant in practice.’
For example, one of the studies was about the dynamics of opinion-forming, in order to understand phenomena such as polarisation. The researchers worked on models that can explain a broader spectrum of opinion dynamics. NETWORKS also ran a project with civil engineers, which recently led to a publication in a renowned academic transport journal.
Mandjes: ‘The research was about routing in a network in which accidents can occur. This model goes back to Dijkstra in a way, but our model allows you to take all kinds of extra effects into account: daily patterns and less predictable events such as accidents. We will be talking to TomTom about this soon.’
Back to Edsger Dijkstra. One of his most famous statements is: ‘It is not the task of the University to offer what society asks for, but to give what society needs.’ A university must respond to what will be needed in our society in the future. This is the only way to stay at the top as a ’network country'. In order to do so, you must constantly develop as a researcher.
Mandjes confirms this: ‘The challenge is to keep up with developments. For example, take the large amounts of data that have become available in recent years. How can you use this data when managing networks? To find out, I will need to learn about areas that are new to me, such as machine learning and other data-driven techniques. There are fantastic opportunities here.’
Annual review Faculty of Science 2021
This interview was also published in the annual review of the University of Amsterdam Faculty of Science. Read our annual review for news and background on teaching and research at the Faculty of Science in 2021, including interviews with lecturers, researchers and students, facts and figures on enrolment and staff news about organisation developments and our valorisation activities.