Bachelor thesis · 2026
Service Blueprinting for an AI supported assistance system in care.
How can the service processes of an AI supported assistance system in residential care be structured so that design recommendations for its integration follow from them.
- Status
- In progress, submission September 2026
- Role
- Author
The case in 60 seconds
- Research question
- How can the service processes of an AI supported assistance system be structured so that design recommendations for its integration follow.
- Method
- Design Science Research, carried out with Service Blueprinting.
- Artefact
- Service blueprints for four typical MAUTZI usage scenarios.
- Result
- Technical and organisational design requirements for integration.
- Context
- Supervised at the junior professorship for Information Systems, University of Göttingen.
- Main limitation
- No regular operation at the time of the study, hence an argumentative evaluation.
Starting point
Care facilities are complex service systems. New technology intervenes in established routines of many people.
Existing systems mostly relieve staff. Resident centred voice assistance is barely researched. That is where the thesis starts.
Method
The thesis follows Design Science Research. The focus is a concrete artefact and the justification of its usefulness.
Service Blueprinting separates frontstage, backstage and support and shows what really happens at each touchpoint.
The artefact
Four scenarios as service blueprints. Starting a conversation, appointment capture, a voice game and a weather request.
Across all scenarios patterns emerge. One visible actor, two processing paths, continuous network dependency, local data.
Result
From those patterns come design requirements, separated into a technical and an organisational side.
One recurring point is the human fallback. MAUTZI complements care, it does not replace it. That matches the German Ethics Council.
Limits and honesty
MAUTZI was not in regular operation at the time of the study, so a trial under real conditions was not possible.
The evaluation is argumentative and analytical. The next step is a real trial in daily operation.
Tools and methods
- Design Science Research
- Service Blueprinting
- Information Systems
Key insight
A system does not become integrable just because it works. Only once the processes around it are visible can you say what the facility and the technology really have to deliver.