It’s all in the data: how data analysts make Bosch’s Dresden fab go round
Lisa-Katharina Reimann on her job at the Dresden semiconductor factory
Semiconductor manufacturing comprises a myriad of complex process steps. So how do you know where exactly a mistake occurred in the case of faulty chips? And is it possible to accurately predict the quality and yield of chips early-on in the process? These are the questions Lisa-Katharina Reimann deals with in her job at our semiconductor fab in Dresden every single day. Her answer: data.
Dresden fab: data-centric from the start
Lisa is a data scientist, analysing fab data for Dresden’s technology department. “We use data to control processes, maintain a stable production, and keep an eye on the product quality of our semiconductors,” she explains. “At the Dresden fab, everything has been data-centric from the start. Even before the first processing happens, we are already collecting data.” This data comes from a multitude of sources, like geometric or optical inspections, electrical tests of chips during production or directly from the manufacturing equipment.
Collaboration is key
Due to the enormous quantity of information collected every day, much of Lisa’s work lies in choosing and developing analytical models and programming machine learning tools to find meaningful patterns: Which parameters point towards optimization potentials in the production process? And how can the continuous analysis of these parameters be automated to make sure that problems are detected before they affect the chip quality?
For Lisa, the key is collaboration with the production experts. “It’s crucial for the data team to work closely with the engineers and product owners in the fab. They know their processes best and can tell us which parameters we need to keep an eye on,” she explains. “And in the end, these experts are the ones who need to be able to work with the results of our analyses and the machine learning models or tools we develop – so a close collaboration also ensures we can support them in the best possible way.”
From maths to microelectronics
Lisa has now been with the data team for more than four years. But if you had asked her during her school years where she would like to work, a semiconductor fab would probably not have been top of her list. “I always liked mathematics for how logical, structured, and plausible everything is – but physics and electronics were a closed book to me at school. The same letters I knew and loved in mathematical equations were suddenly supposed to have a clear meaning in physics, such as units or fixed parameters,” she remembers, laughing. So what brought her to Bosch?
The answer to this question begins at university, where Lisa initially studied business mathematics for her bachelor’s degree – and quickly realized that her passion lay purely in the maths side of her studies. For her master’s, she therefore switched to mathematics at Technical University Dresden – without knowing that her alma mater would end up becoming an important strategic partner for Bosch: Since 2019, Bosch has been cooperating with TU Dresden to welcome students from all over the world to the yearly Dresden Microelectronic Academy.
A degree in MINT teaches you endurance and how to break down and understand difficult topics
During her master’s, it became clear to Lisa that she wanted to apply her knowledge in the industry. “I love seeing how many things in life can be explained with data,” she says. “A degree in MINT teaches you endurance and how to break down and understand difficult topics by yourself. That helped me tremendously to delve into microelectronic topics I didn’t care about or really knew existed back in school.” With Silicon Saxony right around the corner, Bosch became an interesting choice for putting Lisa’s hard-earned knowledge into practice.
Responsibility and appreciation
Even before the Bosch semiconductor fab in Dresden was officially opened, Lisa applied to write her master’s thesis on machine learning and yield prediction there – and quickly decided to stay for good. “The flexibility and appreciation I found at Bosch from the first day convinced me”, she says. “My team trusted me with a lot of responsibility from the start: In my first year as a permanent employee, I was already allowed to become a product owner for one of our product families. This helped me see not just the mathematical but also the technological side of our production.”
In the years since, Lisa has seen the Dresden site grow from a small factory with only 150 employees to one of the most modern semiconductor fab in the world, with more than 10,000 square meters of clean room, and several hundred colleagues. “You can really feel the growth Dresden has gone through in the past few years. But we have only grown closer together as a team the more people we have become,” she says. “Every single day, I am fascinated by all the bright people working at this one fab. I always say that everyone in my team is a unicorn: They’re all uniquely great at something, which helps us solve even the most difficult challenges. We support each other in turning difficulties around, and we always celebrate our successes together.”
Today, Lisa looks back gratefully on the road that brought her to Bosch and Dresden. While her job may sometimes be challenging – “some days, programming seemingly just doesn’t want to work,” she says, laughing – it never gets boring: “It’s very rewarding to see how purely mathematical considerations can lead us to something with a real and helpful impact. And there’s always something new to learn: At Bosch, I have the freedom and opportunity to try new things and push them forward.”
