3D-Printing Masks for COVID
My Mask Movement
3D-printed custom-fit masks for frontline workers during COVID-19.
CORE USER PROBLEM
Traditional mask designs often cater to Caucasian male facial structures. During the COVID-19 PPE shortage, it was crucial for frontline workers and immunocompromised individuals to access masks that fit securely and effectively. My Mask Movement leverages depth-sensing technology and AI to ensure a precise fit for diverse facial structures.
MY ROLE
As the first design hire, I established the brand identity and developed the mobile app prototype. I handled product research, stakeholder communication, design, and prototyping, and coordinated weekly meetings with developers and researchers to drive rapid, inclusive iterations.
RESEARCH & TESTING
We tested the app with doctors and nurses at Stanford Med and Boston University hospitals. Their feedback highlighted key user needs:
Intuitive: Simplify onboarding for less tech-savvy users.
Accessible: Include features like voice commands, text readers, and larger text to accommodate workers wearing bulky gear or glasses.
Transparent: Ensure users feel secure about their data privacy.
DESIGN PRINCIPLES
DESIGN PROCESS
I always begin with "spaghetti sketching"—quick doodles to simplify complex ideas. If a concept isn’t clear in a sketch, it won’t work as a high-fidelity prototype.
Key onboarding decisions:
Simplicity: How do we convey accurate face scan requirements (e.g., good lighting, no glasses, centered face) to a diverse audience?
Solution: We used visual elements like icons and illustrations, making instructions clear across cultures, languages, and age groups.
Result: Onboarding completion rate increased by 15%.
Trust: How do we address user privacy concerns?
Solution: We included a dedicated screen explaining that data is locally stored and never shared, ensuring transparency and building trust.
Key camera capture decisions:
Capture accuracy: The AI to estimate mask size is only as powerful as the data it gets. I consulted our engineer about which variables were most important for an accurate measurement (wide, open eyes; all face angles, a smile) to inform my design flow.
Solution: By moving the camera view from the middle of the screen to the top, users will naturally open up their eyes more because they’re looking directly into the camera on the top of their phone. Arrow animations instructed users on when to move away or towards the camera, and when to tilt their head. We also required the user to take the face scan twice, instead of once.
Results: Sacrificed a 5% drop in completion rate due to having to scan twice for a 20% increase in accuracy.
Accessibility: Users who were required to take off glasses weren’t able to read the instructions.
Solution: Audio instructions were set to default on.
Abandonment: Initially, 1/3 of the users who went through the camera capture would abandon the flow right before they received their results.
Solution: Through user testing, we realized some results could take up to a minute to generate. Users would get impatient while waiting and exit the app. To prevent this, I designed an animation of our elephant “mask-ot” opening up a surprise box with a progress bar. In interviews, users loved the cute animation and the feeling of anticipation for their results.
Results: Abandonment rate decreased by 20%.
Key results page decisions:
Understanding: The National Institute of Health wanted us to display mask sizes using technical terms (e.g., PCA score, Bivariate Panels). However, frontline medical workers preferred simple sizes (small, medium, large).
Solution: We used layman terms on the main screen, with an option to view PCA and Bivariate Panel scores for those interested.
Confidence: Our AI isn’t perfect, and errors or brand size variations could result in incorrect sizing. Inaccurate results led to users firing us.
Solution: To increase user confidence, we allowed adjustments based on detected face nodes, giving users more control over their results.
Results: App deletions decreased by 5%. In interviews, users felt empowered by these controls and appreciated contributing to more inclusive mask sizing through feedback.
PRODUCTION
Initially, we bootstrapped by crowdsourcing 3D printers from the founder’s network and Twitter users. As we gained momentum, we partnered with Stanley Black & Decker to scale production at cost. Our team of ten consisted of two engineers and medical workers as consultants.
In 2024, MyMaskMovement secured a grant from the National Institute for Occupational Safety and Health (NIOSH) to develop a scalable app for frontline workers nationwide. The startup is working towards launching a closed beta with select hospitals by 2027.
Our Partners
Let’s work together.
If you have a project in mind, or would like to chat, shoot me an email at laura.y.gao@gmail.com.