Data and AI-Driven Methods in Engineering – Institute of Mechatronic Systems

Type of course	Lecturer + Excercise (Language: English)
Level	Master
Semester	Summer semester
Credit points	5 CP
Workload	3 contact hours per week
Examination	Written exam (60 min)
Lecturer	Prof. Dr.-Ing. Thomas Seel
Excercise Instructor	M. Sc. Jan-Hendrik Ewering

Course objective

As part of the course "Data and AI-driven Methods in Engineering," students explore innovative methods from the field of artificial intelligence within the context of relevant engineering applications. They learn to tackle complex, real-world problems using appropriate algorithms, including tools from machine learning.

Course content

Introduction and classification of relevant problems and methods
- Core principles of machine learning and artificial intelligence
- Overview of impactful and sustainable engineering use cases
Key cross-disciplinary concepts, including:
- The sim-to-real gap, transfer learning, and domain adaptation
- Hybrid approaches and physics-informed machine learning
- Semi-supervised learning, active learning, incremental learning, and online learning
Explainability, safety, security, reliability, and resilience
Data- and AI-driven methods in simulation and optimization:
- Applying machine learning to solve complex optimization problems
- Using surrogate models for simulation and model order reduction
- Kriging and Gaussian processes in engineering contexts
Data- and AI-driven methods for data analysis and decision-making
Data mining techniques tailored to engineering challenges
- Predictive maintenance and data-driven digital twins
- AI-supported planning, decision-making, and expert systems
Data- and AI-driven methods for physical interaction:
- Bayesian approaches for sensor and information fusion
- AI in motion planning
- Learning-based control of dynamic systems

Materials

The lecture-related materials for the course "Data and AI-driven Methods in Engineering" will be made available for download on Stud.IP throughout the course of the lecture. These include the lecture slides, exercises, exam materials, and supplementary literature.

Important notes

The main programming exercises (each 45 minutes long) take place every two weeks. In the alternating weeks (when no main programming session is scheduled), optional formats are offered to support the course—such as programming office hours or a journal club.