Nina Quante has always been fascinated with quantifying things. It stood to reason that she would turn her passion into a profession. Today, as a data scientist, she supports the Freudenberg Group to identify potential areas for optimization. In determining problems and deriving solutions, she puts the mystery of huge datasets to work.
Quante loves a challenge – especially in her free-time. The 31-year-old enjoys taking part in running events, her sights firmly set on running a half marathon. Not even the idea of a triathlon daunts her. Born in Hanover, Quante had already stepped into unknown territory during her International Economics studies in Tübingen and Göttingen. She not only completed a semester in Spain, but later, also went to Shanghai. “I wanted to experience first-hand what drives China and how the country became what it is.” In 2015, while witnessing a nation amidst enormous upheaval, she could still make out the origins of the country from the cityscape.
I wanted to experience first-hand what drives China and how the country became what it is.
Nina Quante, data scientist
A fascination for numbers and international economics
In addition to internationality, Quante was fascinated with something else during her studies: Data analysis. “My degree largely involved quantitative subjects such as econometrics, statistics and exploratory data analysis. I focused my studies on these subjects very early on, since I have a passion for numbers and what can be derived from them,” says Quante.
Data upon data: As a data scientist, Nina Quante uses data to improve production processes and product quality at Freudenberg. (Photo: Marco Schilling)
Our task is to ascertain correlations which are not at first sight obvious from the data we collect, and to derive decisions.
So Quante’s career path to Data Scientist was already mapped out. But what precisely does a data scientist do? Quante answers: “Our task is to ascertain correlations which are not at first sight obvious from the data we collect, and to derive decisions.” In essence, it means taking a data-based approach to solving problems. At Freudenberg, the focus is on improving product quality, and making development and production processes more efficient.
For Quante, fully understanding her colleagues’ concerns is key. Only then, does she focus her attention on the data collected from sensors.
First, she cleans up the data. Then, she prepares it to optimize the oftentimes sub-optimal data quality. “The more data we have, the easier it is for us to write an algorithm. This allows us to shed light on those factors that affect a result, and which are generally also at the root of the problem,” says Quante.
The data-based approach was entirely new for Freudenberg’s Business Groups. The traditional approach had always been to closely examine the production process and adapt it. Now, digitalization allows data scientists, like Quante, to take a deep dive into production processes which then paves the way for additional value-adding solutions.
Varied sphere of activity
Alongside a cool head, the economist’s job also demands perseverance. Achievement of a goal is an iterative process whereby individual steps are adjusted until the results are informative enough. “Over time, you develop a very good sense of what is right”, Quante says, stressing how important experience is. For her, not all data is the same. It depends on the context in which the data is collected. As Quante points out, major differences also exist from industry to industry.
Over time, you develop a very good sense of what is right
Since 2019, she has used her expertise to support many Freudenberg Business Groups and is delighted to work in such a diverse environment: “During my studies, I had the opportunity to get to know a major automotive supplier. I enjoyed it very much there: So, when I joined Freudenberg, I was looking for a varied job in an international company. And my expectations have been met.”
The solution lies in one gigabyte of data
At Freudenberg, Quante deals with cleaning agents, filters, seals, and nonwovens. Recently, with the latter, she had a challenge on her hands. Freudenberg nonwoven carpets are used in cars. But in rare cases, irregularities can occur during processing of the fleece, which might lead to complaints. Fortunately, Freudenberg had already identified the cause. The nonwoven material is formed by taking melted polymers, processing them into fibers and depositing them onto a collection belt. The fibers mustn’t be allowed to tear, however. When this does occur, it results in irregularities in the fleece. The challenge was to establish what was behind the tear. An unsolved mystery, until Quante, together with another Data Scientist and a Data Engineer, analyzed the process and quality of data recorded over several years. The three scientists made a careful evaluation of a gigabyte of data. They established the cause, presented their results, and suggested solutions to those responsible. “Our counterparts greatly appreciated our findings. But for other colleagues, our results were a surprise since they had expected a different outcome,” says Quante. “When looking for solutions, we aim to provide our colleagues with interim results and illustrate them in a comprehensible way. This not only leads to improved understanding, but also to greater acceptance.”
Freudenberg values the work of its data analysts and recognizes their contribution to the success of the business. Quante’s team consists of mathematicians, economists and mechanical engineers, young and experienced employees. “Our basic interest is always the same. We seek to find hidden correlations from the data.” To do this, the team relies on new technologies. “Sensors have been recording data for many years. But now, thanks to cloud computing and enormous computing power, we can achieve added value from the data.” For data scientists like Nina Quante, this helps to solve various problems. Albeit that they “only” now know why irregularities occur in the nonwovens at Freudenberg and how to prevent them.
Our basic interest is always the same. We seek to find hidden correlations from the data.