Originally, Andreas Büchse wanted to become a farmer. Today, as a statistician in BASF's research department, he helps to ensure that as many insights as possible are gained from experiments. It's about progress, for example, in the biodegradability of coffee capsules, detergents or salad foil in the field. So he still has a close connection to agriculture today. His statistical models not only optimize products that we use every day – they can even avoid experiments on animals or humans.
As a statistician, you will mainly deal with figures, data, and formulas in research at BASF. Isn't that totally dry?
Not for me. Numbers and other data often contain hidden messages and information that you don't recognize at first glance and that are waiting to be discovered. The search for it is great fun for me. This awakens a certain gold rush mood in me. I try to identify patterns and trends. In the best case, these patterns are not trivial, so that others do not discover them and we have an advantage over competitors in research. Another aspect: I like to help, and as a data scientist you often act a bit like a catalyst that helps colleagues in the team to be successful. That's a nice feeling.
How can we imagine your work as a data scientist practically applied in research?
In research, we try to combine different properties of products in the best possible way. The product, for example packaging, should survive the supermarket, but at the end of its use it should be able to be decomposed by microorganisms with as little residue as possible.
Using statistical models, I can predict properties such as the biodegradability of polymers, the raw materials for plastics. To find the perfect recipe for a new product, I can calculate where the intersection is, i.e. the optimal compromise between durability and degradability, price aspects and all sorts of other factors. In this way, I help my colleagues from different fields of work in the development of new polymers – for example, for effective active ingredients in creams or more environmentally friendly detergents.
What specific products have you worked on?
I was involved in the development of compostable coffee capsules made of the bioplastic ecovio®. It's a nice feeling to have played a part in hopefully avoiding a lot of waste. Of course, coffee basically has a bad ecological balance – I also drink too much myself, but then mostly freshly ground from the machine – without capsules.
Another project is about a film based on shark skin for the rotor blades of wind turbines, BASF is working with wind turbine operators to increase the performance of the turbines. For wind turbine operators, the question naturally arises as to how much more electricity they can generate with the film and whether it is worthwhile to equip their turbines with it. Using statistical models, I evaluate the data from tests on sticker-mounted systems and try to determine the potential additional yield. If we can convince all operators worldwide and increase the wind energy produced with a chemical product, that would be amazing.
How else does data science contribute to advancing research and development?
Research at BASF is mostly empirical, that means it is driven by the evaluation of experimental results. As a researcher, I ask nature, so to speak, a question. We then often generate enormous amounts of measurement results and other data. Statistics help to cope with this flood of data. However, statistics also help to plan the experiments well in advance in order to obtain the most informative data possible. In this way, BASF can draw as much knowledge as possible from the experiments with as little effort as possible.
Data science also makes moral sense in research. This is because it can help to reduce the need for clinical studies with humans or animal experiments to be approved for a product.
One example is UV protection for sunscreens, which have so far been tested on humans at great expense. For example, I have supported my colleagues in research in replacing these tests with in vitro methods and mathematical models in the future.
Researchers are naturally more factual and scientifically oriented. What role does your gut feeling play?
Developing statistical models that allow good predictions is also a bit of an art for me. Good statistical data analysis requires experience and definitely a good gut feeling. In the meantime, data is increasingly being evaluated automatically. This also works well to a certain extent, but there are limits. One should never completely leave out the human being with his extensive contextual knowledge and experience. The chemical and physical processes – especially in formulations – can be very complex, and not everything can be captured by our rather simple and mechanistic models. That is why it is important to always remain humble before the experiment – and of course before the expertise and experience of the chemists.
How did it come about that you embarked on a career as an expert in data science after studying agriculture in Göttingen?
Since I don't come directly from a farm myself, I didn't have a good chance of gaining a foothold in practical farming. I was fascinated by applied research and so I pursued an academic career. At the Institute for Sugar Beet Research in Göttingen, I "slipped" into statistical data analysis during my doctorate on the subject of plant breeding, simply because I had masses of data that had to be summarized. Suddenly, it was fun and then I evaluated agricultural experiments for my colleagues at the institute for a few years.
After that, I was self-employed as a consultant for statistics, worked as an assistant professor at the Chair of Bioinformatics at the University of Hohenheim and then started working for BASF in 2007 at the Limburgerhof Agricultural Center in the field of data management. More than 10 years ago, my research heart finally drew me to BASF's central research department, where I have been conducting very application-oriented research as a data scientist ever since. This is the ideal position for me.
What do you say to young people who lose heart about the future?
We would have to radically change our behavior to stop the climate crisis. But we are obviously not ready for that, so we have to try to minimize the harmful side effects of our actions through technical innovations. We are currently in a critical phase, and in order to achieve something in the short term, research in industry must make a difference – be it in biodegradability or improved efficiency of processes. In my job, I am in contact with many different research areas and see how things are moving in the right direction with combined forces in all areas.
Join the team.
- Become a smart scientist yourself? The world needs innovation. We need you.Explore open positions.