Oops, We Did It Again! – Simulation of problem situations in development processes using agentic AI

The future of interdisciplinary development needs you!

The development of new high-tech systems requires more than just one discipline: only with the help of developers from mechanical engineering, electrical engineering, mechatronics, software engineering, and many other domains can we create the systems of tomorrow. This becomes particularly exciting for cyber-physical systems (CPS), which connect the digital space with the physical world. Do you want to change our understanding of CPS? Then support the SFB Convide with your work!

What we are looking for

Consistency is a fundamental prerequisite for collaborative work. By consistency, we mean that different development artifacts (e.g., software code, a CAD model, or even a physical prototype on the test bench) do not contradict each other and represent the same system. Numerous tools promise exactly that, but do not yet implement it sufficiently. Many inconsistencies arise in development processes—different cycles, development speeds, and team boundaries lead to communication difficulties and ultimately to errors in product development. We want to исследate this phenomenon—and you can contribute!

Your task

The development of new process models requires validation—and this is difficult to achieve in practice within conventional development environments. We are looking for motivated students who, with the help of generative AI agents, build development processes and the necessary information within them as a simulation environment, and validate conventional process models as well as novel consistency-enabled processes. This includes the implementation of the agentic workflow, the reproduction of known problem situations, and the generation of new process datasets

Am I suitable?

If you have a technical background (e.g., mechanical engineering, electrical engineering, mechatronics, computer science, industrial engineering, ...) and are interested in the development process of complex cyber-physical systems, you already meet most of the requirements! Prior knowledge of using agentic AI is helpful but not required. If you are looking for supervision for a cooperatively supervised thesis in industry, feel free to contact us.

Your benefits

• Work on an interdisciplinary topic between mechanical engineering, electrical engineering, software, and industrial engineering
• Learning scientific working methods
• Intensive 1:1 coaching with a research associate at IPEK
• Scientific publication of the results is intended

Interested? Feel free to write me an email at thomas.voelk∂kit.edu!