From an information technology perspective, modern production systems, such as automated pharmaceutical production plants, are complex communication and interaction networks. A digital twin of this system must not only map individual components, but also precisely capture their dynamic relationships and interactions. Without sophisticated referencing methods, it would be impossible to answer the following questions:

• How do component disruptions affect the production chain?
• Which production steps are affected by component changes?
• How can optimization potentials be identified in real time?

Effective references enable semantically rich modeling that not only depicts states, but also makes causalities, dependencies and development paths transparent.

GOALS

The master's thesis critically examines how such referencing methods affect the information integrity and dynamics of models. The focus is in particular on:

• Semantic losses due to unidirectional references
• Limited traceability of system states
• Challenges in the dynamic modeling of dependencies

The aim is to develop a conceptual framework that optimizes bidirectional and context-sensitive referencing strategies for digital twins. The developed concept will then be implemented as a prototype and evaluated in a real use case.

HELPFUL PRIOR KNOWLEDGE

• Great interest in digital twins and information models
• Programming skills (Go/Python/Rust/...) and an enjoyment of software development
• Contents of the lecture Cyber-Physical Production Systems