Advanced Seminar on Safe Cyber-Physical Systems

Lecturer (assistant)
Duration3 SWS
TermWintersemester 2022/23
Language of instructionEnglish
Position within curriculaSee TUMonline
DatesSee TUMonline


Admission information


After participating in the module, students will be able to understand the basics of cyber-physical systems (CPS), analyze CPS described in the literature, and subsequently evaluate them. The students will also get a deeper understanding of safety in CPS and get an overview of current problems and research topics in the field of CPS. The students further will learn methods for literature research, diversification of sources, and apply scientific methods for literature research.


Basics of CPS, Analysis of CPS, Safe CPS, Methods of Literature Research and Diversification of Sources. This course will introduce advanced topics for safe cyber-physical systems (CPS). Due to the complexity of today's cyber-physical systems, the range of topics covers multiple aspects of computer engineering and control theory, for example, from safeguards for AI-based control algorithms, over robustness against various attacks on network communication, to real-time guarantees in heterogeneous multiprocessor systems-on-a-chip (MPSoC) architectures with embedded GPUs and FPGAs. The students will learn the basics of CPS design and analysis and get a good understanding of the safety problems in today's CPS. The topics covered in this course include a wide range of aspects from existing research, such as challenges of flexible manufacturing, the integration of machine learning and deep neural networks, safety of control algorithms in general and especially when integrating AI-based control mechanisms, energy consumption concerns in mobile CPS, error isolation and fault tolerance, cyber-attacks, real-time scheduling on multiprocessor systems, predictable operating systems and hypervisors, partitioning of the memory subsystem. Furthermore, it is possible for students to suggest their own topics within the seminar's thematic scope.


There are no compulsory pre-requisites but note that the course is taught on a Master degree level. As this course is open for students from both informatics and mechanical engineering, a wide range of interdisciplinary areas are covered, so an interest in either control theory, machine learning, computer engineering, real-time systems, or operating systems is helpful for research in specific topics.

Teaching and learning methods

In the first sessions, a set of current research topics will be presented by the lecturer. Students then select a topic of choice and start literature research in their area of interest. In subsequent sessions, the students present their progress, and the lecturer discusses the results and guides further research. The students prepare and write a scientific report on their topic and they hold a presentation at the end of the semester. In order to optimally support the students, a small group size is aimed for this course. Due to the small number of participants, the teaching method consists of individual professional and methodical discussions between the lecturer and students as well as presentations by the lecturer and students in front of the group followed by open discussions.


The learning outcomes of the students are checked by means of a report (including a presentation). At the end of the seminar, students submit a written report of their literature review, which is graded by the lecturer (70%). In addition, the students must prepare a presentation that will be presented to the students and lecturer and will be graded (30%).

Recommended literature

K. Kim and P. Kumar, “Cyber–Physical Systems: A Perspective at the Centennial”, Proceedings of the IEEE, Vol. 100, May 2012. DOI: 10.1109/JPROC.2012.2189792