When a change is made in an industrial Digital Twin – for instance, a property in the type description of a device is deleted – the question arises: Which concrete devices (instances) are affected and need to be adapted? A naive answer: all instances of this type. But this is often too far-reaching. If an optional property is deleted, only the instances that actually use this property are affected. The challenge: one must distinguish between two things: (1) Structural Reachability – all instances that can technically be "reached" from the change because connections exist. (2) Semantic Affectedness – the true subset thereof that actually needs to be adapted because the change influences their meaning or function. This master thesis develops rules to systematically make this distinction and to make change management in Digital Twins more efficient. 

GOALS:

This thesis combines concepts from graph theory (networks of linked information), model analysis, and change management. Industrial Digital Twins can be modeled as networks in which type descriptions and their instances are connected through different types of relationships. Semantic affectedness requires additional knowledge: What is the meaning of the change? How do instances use the changed information? 
The thesis develops rules (heuristics) that decide, based on the change type and context, whether a reachable instance is actually affected. Basic knowledge in graph theory and information modeling is required.

HELPFUL PRIOR KNOWLEDGE:

• Information Meta-Modeling (AAS, RAMI4.0)
• Software-Architectures 
• CPM / CPPS