From data complexity to reliable and intelligent sensor solutions

I work where sensor technology and data science meet. I aim to help colleagues make better and faster decisions by translating numbers, signals, and raw data into comprehensive information.

My role is to advance data‑driven development approaches for MEMS sensor systems, from defining suitable algorithms and AI methods to ensuring their robust integration into series products. The team I lead is dedicated to data science and digitalization. Additionally, I contribute to strategic decisions on how data‑centric development can strengthen our sensor portfolio, scale efficiently across applications, and position Bosch as a center of excellence for data‑driven innovation.

Sensor development has evolved from a predominantly hardware‑driven discipline to a strongly data‑ and software‑driven field. Literally no decision about sensor concepts and designs is made without extensive data campaigns that simulate and characterize performance and quality.

Today, sensor performance is no longer defined solely by the sensor element itself, but by how intelligently data is processed, interpreted, and combined with algorithms and AI. At the same time, customer expectations regarding accuracy, robustness, and fast time to market have grown and changed in the last ten years. As systems become more complex, data plays the key role in making faster decisions and securing high quality standards, especially in the innovation and development phases where the concept for mass production are set with only a few samples.

I mainly decide how data and algorithms are applied to solve concrete engineering challenges. This includes selecting the appropriate analytical or AI approaches for specific applications, deciding where data‑driven methods generate clear customer benefit, and setting priorities when technical complexity or trade‑offs arise. In case of problems, I ensure they are handled by experts with the right skills so that innovative concepts can be translated into robust and scalable sensor solutions.

What makes my job unique is the combination of deep technical work and continuous learning. I am fascinated by the idea of translating human senses into technical systems and by the clarity that data brings to complex physical behavior. My strongest motivation comes from working with a highly skilled, interdisciplinary team that shares my curiosity and ambition. Together, we continuously challenge existing assumptions and push sensor technology forward through data‑driven insight.

Customers benefit from sensor solutions that are developed with a strong focus on data‑based objectivity and real‑world performance. By combining sensor expertise with advanced algorithms and data analysis, we create robust systems that operate precisely and reliably under diverse conditions. This allows our customers to build their applications on a solid, transparent information basis and to make confident system‑level decisions.

My work contributes by transforming complex data into clear, actionable insights that directly influence product and system development. These insights enhance the design process, help optimize sensor behavior, and reduce uncertainty in critical development phases. As a result, customer projects benefit from improved performance, higher reliability, and solutions that integrate smoothly into larger system architectures.

With the rising popularity of AI and machine learning in academics since 2014, many powerful methods and models were developed that help to solve even most complex problems. By the trend of GenAI, large language models (LLMs) and agentic workflows, those methods become available now to every person around the world by Copilot functions. Sometimes assessments are on pair between experienced colleagues with high academic education compared to LLM answers, e.g., in defect image assessments. We have to understand this as a strategic advantage as well as the toughest competition.

Much of the manual work that we do now can and will be automated in future. As a result, we see real efficiency improvements in our daily work – on one hand. The brightest side on this for me however is, that once proficient enough with new tooling, colleagues are fascinated by the potential to improve their daily work life, which I observe as an enormous boost in their motivation by getting rid of the boring stuff.

Further, much of our work that we do now will be automated – both for data scientists as well as for regular engineers.

No single person can address the complexity of today’s systems alone. And neither can innovative ideas come to life solitarily. Collaboration with the most profound experts is essential for both the sensor and the AI community.

My past roles helped me to build strong and personal connections in the whole semiconductor MEMS community – be it for automotive or consumer electronics. My experience ranges from early innovation phases where concepts sometimes only exist on paper to large-scale manufacturing lines more than ten meters long and producing millions of sensors.

My current role brings connections to data and AI teams across Bosch to exchange methods and competences. Furthermore, as a council member, I help our Bosch Research field and my colleagues worldwide adopt AI techniques to make a significant contribution to the scientific community.