Description

Problem: 

Data-driven Digital Twins often autonomously mirror the characteristics, behavior, and attributes of a real-world entity and propose actions. However, in some Digital Twin systems, humans may still need to intervene or can significantly enhance system operation through domain knowledge and cognitive or physical capabilities. To solve this problem, a structured understanding of when and to what extent humans could intervene or support Digital Twin operations would be necessary.

Goal: 

The goal of this thesis is to identify scenarios in which human support and intervention can improve the operation of data-driven Digital Twin systems and to design suitable intervention techniques.
Example scenarios include data validation, model development and validation in risk-sensitive environments, or use cases with particularly high cognitive workload. Possible intervention techniques include manual override mechanisms, event log filtering, or feedback mechanisms for improving Digital Twin operation.
Based on this, the student will conceptually map identified support scenarios to interaction and intervention techniques from related technological domains, such as Human-Computer Interaction, decision support systems, or autonomous systems, and demonstrate selected concepts in a small-scale Digital Twin example.
The outcome of the thesis is a conceptual framework for mapping human support scenarios to suitable intervention techniques and interaction strategies, as well as an initial Digital Twin prototype implementation.
 

Required Skills and Knowledge: 

• Basic knowledge of simulation systems and Digital Twins.

• Programming skills (Python beneficial).

• Interest in system design and human interaction.

• Analytical thinking.