Description

Problem

Simulation-based Digital Twins (DTs) are widely used for monitoring, prediction, and decision-making in complex systems such as manufacturing. While the underlying models are often transparent in principle, the overall DT system is not. Users typically lack visibility into key aspects of the DT components: where the data comes from, how it is processed, what assumptions shape the model, what validation actually demonstrates, and how to interpret uncertain or stochastic outputs. In practice, most DT interfaces focus on what is happening — system state, alerts — rather than why it is happening or how the results should be trusted. As a result, there is little guidance on how to design interfaces that make these aspects understandable and actionable for different stakeholders.

Goal

This thesis will design and evaluate low-fidelity prototypes that make explainability concrete across the main components of a Digital Twin: Data Collection, Data Validation, Knowledge Extraction, Model Development, Model Validation, and Analysis. The work follows a Design Science Research approach:

1. Identify user needs through literature and stakeholder input (e.g., interviews or workshops)
2. Develop interface prototypes (e.g., in Figma or Balsamiq)
3. Evaluate them using structured methods (e.g., think-aloud studies, cognitive walkthroughs, or expert review)

Research Questions.

1. What explainability needs do different DT stakeholders (e.g., developers, operators, decision-makers) have across the DT components?
2. How can these needs be translated into concrete interface designs, and which patterns emerge across contexts?
3. How do these designs affect users' understanding of DT behavior, as assessed through structured evaluation?

Required Skills and Knowledge

• UI/UX prototyping (e.g., Figma)
• Basic qualitative research methods (e.g., interviews, think-aloud)
• Ability to communicate with both technical and non-technical stakeholders
• Interest in Digital Twins and simulation modelling
• Visual design experience is a plus, not a requirement