Digital Deep Dive: Data Science @ the Digital Hub

Munich, September 13, 2023 – Our Data Scientists are an important part of our in-house development team, because they know how to turn data into gold. By integrating Data Science into the development process, patterns, trends, and correlations can be uncovered to make data-driven decisions for an overall user experience. In the fourth edition of “Digital Deep Dive”, we discover the power of data. Let’s deep dive into Data Science with our expert Pit Vanhoefer, Team Lead Data Science.

Let’s put the spotlight on Data Science with our expert Pit Vanhoefer, Team Lead Data Science, who joined the Digital Hub in 2019.

Let’s directly jump into the topic. What exactly is Data Science?

Pit: You can find Data Science in many areas of our life. Netflix for example combines data on how users interact with the platform including what users watch, how long they watch and what they search for. They collect data about the preferred content, actors, directors, and more. This data is used to train algorithm to recommend new content to the customers, based on what the user liked and watched so far and hence creating a personalized user experience.

Described in a more general way, Data Science combines mathematics and computer science and requires a good business understanding. It feeds upon data and needs a good infrastructure to show full effect. Machine learning introduced a change of paradigm, now using a computer to define the rules to be applied on data to provide actionable insights for decision-making.

And, how does your team contribute to our digital transformation?

Our team is involved in a variety of digital initiatives within Oerlikon. Our aim is to identify areas where Data Science can have a significant impact, and therefore we explore a variety of ideas in different areas, such as R&D, Operations or Sales. We use a good mix of algorithms; some are already widely used in the industry, for example from the field of computer vision, but also others from more recent developments, such as language understanding.

One example in the field of advanced computer vision is the analysis of images of the powder bed in our 3D-printing machines during the printing process. This automated cutting approach makes it possible to detect even small errors in real-time, ensuring effective quality control. An example for using language understanding algorithms is the automatic processing of incoming order documents, where we digitize orders from paper. This simplifies the transmission of order details to our ERP system.

In addition, we are involved in fundamental work like designing our data landscape of the future – from infrastructure to more valuable data assets and data governance. A comprehensive Data Strategy, with regular analyses and evaluations, is key for us. Our objective is to harmonize data from e.g., production and quality control, enabling data-driven strategic decision-making.

How are these projects distributed in your team?

In our team we have five Data Scientists and one Data Engineer responsible for our projects. We have a broad range of experience and currently focus on visual analytics, advanced text processing, and predictive analytics. Each team member works on several projects in parallel, collaborating strongly with our colleagues from the other teams.

Are there any specific technologies or tools that your team is using to support digital transformation?

It is important for us to keep up to date with the latest development in the field of Data Science and to continuously extend our skills with new algorithms, tools, and technologies. We use high-quality, open-source machine learning tools to build our own solution or get existing solutions from the market. Using pre-trained models and adapting them to our tasks allows us to benefit from their capabilities and performance.

As part of the development team, how does your team collaborate with others to drive digital transformation?

Our development team complements each other with expertise in Frontend and Backend Development as well as Cloud Architecture to provide end-to-end solutions. In order to be strongly aligned with the business and understand its needs, our colleagues from the program management play an important role, as well as the know-how and experience of our subject matter experts from the business.

Are there any challenges your team is facing?

The challenge is to find impactful use cases that add value to our customers and match the data we have. Making a real impact requires thorough examination of an often-complex reality. In Data Science we also face the challenge of so-called data silos, where large amounts of data exist, but the data cannot always be linked together in a meaningful way. This is why a comprehensive data strategy to harmonize our data is so important for us.

What projects will your team focus on in the future?

We aim to improve the sales and finance data engagement, exploring untapped potential in areas like forecasting trends and customer behavior, such as customer churn. The recent impressive achievements with large language models are of course very interesting for us, too. We explore how to use generative AI for customers´ and employees´ interactions (e.g., chatbots), smart knowledge management and more.

Pit Vanhoefer in interview with Nazanin Khorram.

Digital Deep Dive: Cloud Architecture @ the Digital Hub

Munich, August 16, 2023 - Cloud Architecture provides the basic structure for maximizing the benefits of cloud computing, such as flexibility, se

Digital Deep Dive: Backend Development @ the Digital Hub

As technology continues to advance, the Backend Development team is making a significant contribution to Oerlikon’s digital transformation by ens