The Dutch water sector is facing new challenges. Rapid climate change is reducing the predictive power of our historical datasets. To prepare for a future with increasing drought, extreme showers and a rising sea level, we need innovative calculation methods and therefore more AI experts. We interviewed Mark Roest, director at Vortech, where they develop software for various social issues, about this.
Mark was a co-author of the DigiShape white paper that led to the new “Water and Climate” working group at the Dutch AI Coalition. Through this connection, he hopes more AI experts and mathematicians will join the water sector.
Mark, tell us a little about yourself and Vortech
“At Vortech we work with about 35 people to develop demanding applications for predictions, simulations and other complex calculations. Our customers come from the water sector, but also from other fields such as energy and environmental management. We believe it is important to provide not only mathematical expertise, but also knowledge of the issue at hand. This way we are able to ask the right questions to experts and can write software that fits the question well. One project I am proud of is Hydra-Ring. This is a calculation program that Rijkswaterstaat uses to determine the strength of dikes. We have accelerated a number of intensive calculation processes, as a result of which the program is now able to perform calculations that were not possible before.”
Why do you think more AI and Machine Learning is needed in the water sector?
“Because with the help of Machine Learning and AI we can perform more and different computational work than was possible until now.In the Netherlands, we are still at the forefront of knowledge about water, but with the increasing impact of climate change, we are facing new challenges for which we have no ready-made solution.Consider the “water bomb” that occurred in Limburg a few years ago.The problem was not only the heavy rainfall, but also the fact that the shower remained in one place for a long time.Local effects such as these are often difficult to predict with traditional models.Another example is the behavior of people during a calamity.With traditional models you can’t do very much with that but with machine learning you might be able to get some grip on that.”
What makes working in the water sector interesting for people with mathematical backgrounds?
“The social relevance. I myself studied physics and went on to do a PhD in mathematics.I can imagine that people with such a background quickly choose a career in finance, or the commercial world, because so much more is already possible there.You can’t apply for a mortgage or click on a video without a world of Machine Learning behind it.In the physical domain, that’s a lot more difficult, because the data is scarcer and less structured.But that’s precisely what makes this work so much fun for me.From a complex puzzle of datasets and interests, I create new possibilities that make the world a little better. Think about increasing water safety and countering damage to nature and agriculture caused by droughts.”
What will the NL AIC’s new ‘Water and Climate’ working group do?
“That has yet to crystallize, but we already have two initiatives in mind. First, we are going to set up a living lab in which we will compile all the data collected during the “water bomb,” including weather forecasts, precipitation measurements and groundwater levels. Together with experts, we want to thoroughly analyze this data: can we use new AI methods to see such a disaster, or the likelihood thereof, coming earlier and predict better than with current methods?In addition, we will brainstorm creatively about the possibilities of generative AI for the water sector.”
