Overview
Insipiration
As designers, we face unique anxiety... How can we navigate it? How could we maintain our mental wellbeing?
This project started from my design studies at Delft University of Technology. As an international student, adapting to a completely different culture and education system in a country is never easy. During my studies, I experienced many challenges, which caused me strong anxiety for a long time. This experience sparked my interest in mental health issues. Therefore, I decided to use this experience as a starting point for my design.
Solution
Lumi - An Al reflection tool for designer (students)
Lumi is a mental health support tool designed for design students’ mental well-being. It aims to help and coach on how to cope with anxiety during the design process and generate AI-enhanced personal insights throughout the design study.
Quite video introduction in case you don't want to read:)
Design process
6
User interviews & Diaries
10+
Brainstorms + Concepts
4
Interface iterations
10
User evaluations
Research
Research background
Conversational AI was widely used for personalized mental conversation
I limited my design solutions into the AI because the technology has been widely used in delivering mental health support. AI chatbot could become a personalised “emotional coach” for design students, helping them deal with anxiety, regulate the emotions.
Interview process
To build design resilience, we need to know what the typical anxieties or challenges are faced by design students.
I adopted a combination of interviews and material-based reflections to understand:
- Typical anxiety faced by design students
- Their inner psychological journey when experiencing anxiety and the ways and mindsets they use to overcome it.
- Their preference of AI chatbots to cope with anxiety.
left: Anxiety self-assessment survey; Right: Anxiety reflection sheet
Mental journey mapping
The journey showed anxiety is usually caused by unhelpful thinking & behavior patterns, same for design students.
I asked the participants to recall one project or a scenario that they experience anxiety to know the cognitive behavior patterns behind anxiety from the psychological perspect.
Problems
Doing a design project is sometimes a roller coaster journey. How can we do and learn from it?
Through observation, interviews, and user research, I analyzed the psychological journey of design students during project work and summarized the main problems they encounter:
So, how could conversational AI help during the design process?
I also analysed how the AI chatbot can help users address specific anxiety issues throughout the process and guide them toward reflection. There are two situations:
Two situations of facing anxiety
Design
Design Solution
Lumi - an AI companion for designers
Lumi is a mental health support tool designed for design students’ mental well-being. It aims to help and coach on how to cope with anxiety during the design process and generate personal insights throughout the design study.
Step 1
Create the project
Create a design project that you’re doing now and add the milestones for system to send a mental health tracking reminder.
Step 2
Project feeling log in & AI chatting
- Indicate the current feeling about the project
- Choose on what factor(s) trigger the feeling
- (Optional) Chat with Lumi about the feeling.
- Log in the feeling in the project journey
*The dialogue was pre-trained
Step 3
Check the project journey
It will show an emotional journey of the project on the timeline with AI insights and triggers lead to that feelings
Step 4
End-of-project reflection (AI-generated)
Journey overview
Analysis on project feelings
Analysis on triggers behind feelings
Lumi’s recommendation
Bonus!
Overall personal insights (AI-generated)
The students can ask Lumi about the insights of all the projects. Those questions link to the understanding of themselves as designers, like:
- “What kind of designer am I?”,
- “What aspects can I improve?”
- “What’re my strength?”
- “Advice on my future career.”
- ……
Future development
A plugin in Figma
This could lower the threshold for users to use, as they don’t need to open a new website. Figma or Miro is frequently used by designers in daily lives. They are more likely to use the tool to write their reflections. And it can be easily linked to their design process.
Prompting with Voiceflow and ChatGPTs
The prototype doesn’t integrate the OpenAI API. In order to test the result of conversations, I used Voiceflow to simulate conversational steps (indicated before), and finally, I decided to use ChatGPT to simulate the chatting (design anxiety coping + design reflection) function and compared the emotions before and after using the prototypes.
* I iterated on several versions of prompts for its best performance.
Conversational flow in Voiceflow
Result
User test
Students expressed 80% willingness to use Lumi for the design study
Lumi is perceived as visually engaging and attractive, but the introduction process requires refinement. A clearer and more interactive introduction could help participants quickly understand how the tool works.
Qualitative results
also shows high score in Attractiveness, efficiency, and novelty
Participants rated Lumi particularly high on Attractiveness, Efficiency and Novelty, with a particular high score on feasibility, creativity and high value.
