Data-driven Engineering
This profile focuses on data-driven methods for monitoring, modeling, and optimizing engineering systems, including predictive maintenance, scientific machine learning, and digital twins.
Predictive Maintenance: Knowing When to Act
Machines give warning signals long before they break down. Data-driven models read these signals from sensors on bearings, gearboxes, or turbine blades, estimate how much life is left, and flag when it is time to act. Not too early, which wastes money. Not too late, which wastes even more. The resulting longer component lifetimes and fewer unplanned replacements directly reduce material consumption and waste, contributing to a more sustainable practice.
Scientific Machine Learning: Test Less, Learn More
Every experiment costs time and money. Scientific machine learning combines physical knowledge with data to build models that predict the behaviour of complex structures with minimal experimental or computational effort, and that tell you how confident they are in their own predictions. Fewer physical tests also means less energy, less material, and a smaller environmental footprint throughout the development process, and helps creating better products faster.
Digital Twins: Your Component, Simulated in Real Time
Physics-based simulations are accurate but far too slow for real-time use. A digital twin is a data-trained copy of a real component that delivers the same results in milliseconds, continuously updated with live sensor readings from the machine it mirrors. In this way, we can let machines operate closer to their true limits, rather than conservative worst-case assumptions, which saves energy and extends service life.
The profile consists of the module catalog C and the module catalog D of the profile. By choosing this profile it is possible to absolve the Master´s program in English. Detailed descriptions of the selectable modules can be found in the module handbook of the Master’s program in Mechanical Engineering.
Competencies and Career Fields
The newly developed Data-driven Engineering profile introduces students to modern computational and statistical methods that connect data with physics-based simulation models to create integrated digital engineering workflows. Key topics include machine learning, optimization, model updating, uncertainty quantification, process control, and digital twin technologies.
Students learn how to extract meaningful insights from experimental and operational data and integrate these insights into simulation and decision-making processes. Typical application areas include structural health monitoring, predictive maintenance, system identification, control engineering, and mechanical reliability analysis.
By completing this profile, students gain the skills needed to design and implement data-driven engineering solutions in both research and industrial environments.