User-centered Assistance and Safety System

Category: Tablet App Time: 6 Months 2020


Development of a user-centred Assistance and Safety System for caretakers and patients with dementia. Scope covered research, concept and prototyping in an interdisciplinary university project, followed by an independent extension into UI design, a design system and an AI-assisted prototype.


Results

Results

The research project delivered a summary of findings into the day-to-day work of professional caretakers and a first version of the app prototype.

After the project concluded, I continued independently: designing wireframes, building a design system and developing an AI-assisted prototype in Figma and Claude Code using Visual Studio.

Because a usability test with caretakers wasn't possible (because of the Covid pandemic) and the later extension was self-directed, there are no live user metrics for either phase.

Role in Project

Role in Project

Research and prototyping, in a team of three M.Sc. Human Factors students during the project semester. Tasks were split evenly across the team: I led parts of the focus group, took part in the ideation sessions for concept creation, created the wireframes in Axure RP, led multiple expert review sessions and contributed to the data analysis.

After the project ended, I continued alone, working on the concept, UI design, and the AI-assisted prototype and design system as a personal project.

Problem Statement

Problem Statement & Goals

Develop technological support for professional caretakers specialised in caring for clients with dementia by using tablets with a specialised app. Most importantly, the project included determining which technological support would really help them in order to avoid additional workload.

The goal of this project was finding a solution to help the caretakers by simplifying their care records, which they currently have to complete on a smartphone app and additionally as handwritten documentation. Additionally, the system shall show predictions about the patients’ health status and potential risks.

Opportunity

Problem Statement & Goals

From the focus group findings, we defined the app's key functionalities together, prioritising what caretakers actually needed. The project's scope stopped at a first prototype and review. After it ended, I decided to take the opportunity to use the expert review feedback to design an improved concept to deepen my own concept and visual design skills. From there, I decided to use a AI-assisted workflow to create a functional prototype including a design system to build my own skills in AI-assisted design and prototyping.

Users

Users

Professional caretakers specialised in caring for clients with dementia

Design Process

Design Process

Literature research

Literature research was carried out about the following subjects: UIs for caretakers, technological support for retirement homes and hospitals, symptoms of dementia

Focus Group
As a second step, a Focus Group at a retirement home was carried out in order to better understand the needs of the professional caretakers.

The focus group included:

  • Creating personas

  • Creation of a daily calendar to provide an overview of the caretakers‘ tasks

  • Brainstorming: Identifying essential support functions required for an in-home care interface


Literature research

Literature research was carried out about the following subjects: UIs for caretakers, technological support for retirement homes and hospitals, symptoms of dementia

Focus Group
As a second step, a Focus Group at a retirement home was carried out in order to better understand the needs of the professional caretakers.

The focus group included:

  • Creating personas

  • Creation of a daily calendar to provide an overview of the caretakers‘ tasks

  • Brainstorming: Identifying essential support functions required for an in-home care interface


Brainstorming Persona Focus group
Brainstorming board Focus group

Paperprototyping & Wireframing
Considering the results of the Focus Group, the key functionalities were defined. A prototype was created using the methods paper prototyping and wireframing (Axure RP). Since the UI is specific to the use on a tablet, special attention was given to the element size.

Wireframes for UI

Expert Review
An online Expert Review with usability experts was carried out in order to gain further insights for the implementation of the app. (Due to the pandemic, a usability test with the caregivers was not possible).

Feedback of the expert review included:

  • Seperate menu items related to the patient and the caretaker to make the navigation more intuitive

  • Display critical alerts and warnings in a dedicated section on top of the home screen (heart rate anomalies, emergency signals, or important notes from the care protocol)

  • Include biofeedback data on the home screen to give caregivers an immediate overview of the patient’s current condition

  • Present the information gathered by the camera system in a more visual format


Expert Review
An online Expert Review with usability experts was carried out in order to gain further insights for the implementation of the app. (Due to the pandemic, a usability test with the caregivers was not possible).

Feedback of the expert review included:

  • Seperate menu items related to the patient and the caretaker to make the navigation more intuitive

  • Display critical alerts and warnings in a dedicated section on top of the home screen (heart rate anomalies, emergency signals, or important notes from the care protocol)

  • Include biofeedback data on the home screen to give caregivers an immediate overview of the patient’s current condition

  • Present the information gathered by the camera system in a more visual format


UI Design
Although UI design wasn’t originally part of the project scope, I decided to use the expert feedback for creating an improved concept and UI design after the project was completed. This allowed me to deepen my concept and visual design skills.

AI Prototype and Design System
This step included researching best practices for AI-assisted UI prototyping (thinking, frameworks, checkpoints, debugging), writing project specifications with milestones, selecting a tech stack with AI-generated and evaluated options, building a design system to guide the visual design, recreating all existing wireframes in VS Code using Claude, and prompting new functionality into the prototype. I reviewed each AI-generated iteration for accessibility issues, inconsistent typography and iconography, and responsive design gaps, refining the prompts until the result held up, and adding the new elements back into the design system.

AI Prototype and Design System
This step included researching best practices for AI-assisted UI prototyping (thinking, frameworks, checkpoints, debugging), writing project specifications with milestones, selecting a tech stack with AI-generated and evaluated options, building a design system to guide the visual design, recreating all existing wireframes in VS Code using Claude, and prompting new functionality into the prototype. I reviewed each AI-generated iteration for accessibility issues, inconsistent typography and iconography, and responsive design gaps, refining the prompts until the result held up, and adding the new elements back into the design system.

Video example: User is looking for information on the patient and todays activities



Video example: User is looking for information on the patient and todays activities



Conclusion

Con
clusion

This project took important steps toward simplifying documentation for caretakers working with dementia patients. For a project this complex, the next step would be testing the app with caretakers again and iterating on the documentation functionality until it genuinely fits their everyday workflow which was prevented by the pandemic.

Working with an AI-assisted workflow also taught me to stay critical of what gets generated: without actively reviewing for accessibility, typography and responsiveness, inconsistencies slip through easily. Catching and correcting these became as much a part of the process as the prompting itself.

Location

Copenhagen, Denmark

Munich, Germany