Problem Statement

How might we help encourage people to wear masks when doing outdoor exercises?


Smart MaskSolution.png

Design Process

Design ProcessSmart Mask.png

Research Question


What is the biggest concern for the mask-wearing experience?


Are there any existing masks to improve breathability? How do they work?


What respiratory data can we collect and do users concern about? 

Check the Full Research Report here: 

Competitive Analysis



To properly scope our topic and confirm or deny our assumptions, we designed the first survey to ask participants about their past experience of wearing a mask.
We run a pilot survey first and received 10 responses. Based on that we revised our survey questions and received 98 responses for the official survey. Check the report here.


80% of the participants thought that “physical work or exercise” was a significant contribution to discomfort when wearing a mask.

The most prioritized concern is “Fogging of eyewear”, of which the average rating is 3.4. The second biggest concern is “It is humid and hot inside the mask”, of which the average rating is 3.0676.
And then the following several important problems are “Mask becomes unsanitary quickly”.

The survey confirmed our hypothesis:

  1. The fogging problem and discomfort caused by high humidity is the most common concern.

  2. And the outdoor exercise is the most common scenario.

However, communication-related issues averaged low scores on the problem rating scale.


We found that some problem descriptions were too broad, and the interviewee did not clearly know what we meant. For example, in the question about “Irritation behind the ears from mask loops”, we wanted to know if people felt pain or irritation after they wore the mask. However, major respondents rated this as minor problems despite the comments in the end: “adjustable ear ties”, “part of the ears be more comfortable”, and so on. we should avoid using too broad words like "Irritation". 



45 responses 

In order to explore the possibility of the data we collected, we designed a mix of a quantitative and qualitative survey. Check the survey question here.

The survey was distributed from Prolific and participants screened for people who exercise regularly.  Here are the survey results and analysis.

Target users

95% of participants doing exercise more than 60 mins per week.
83.3% of participants use health tracking devices/apps now. 
64% of them check the apps more than 3 times per week.

37 of 45 have never tracked the respiratory data.

Preferred features

75% of the participants used “Tracking behaviors.”
Besides that, people prefer historical data & personalized suggestions.

50% of participants view "Tracking behaviors" as very useful ( 5 for 1-5)

What're the concerns for health wearable

Users are most concerned about accuracy, privacy, and battery life.
Few users think the summarized information is useless. 
Some users need reminders.

chart (1).png
What data users want to know

Users are most interested in breathing data, mask usage data, and sensor data in order.
For sensor data, they are most concerned about humidity.


Most interested in respiratory data (current breathing rate)​

They care about battery life and usage time for the current mask.

They prefer tracking behaviors most.

Privacy is the broadest user concern.

Importance-difficulty matrix

By mapping those user requirements we gathered from the survey into the importance-difficulty matrix, we found top prioritized requirements are those with high importance and low difficulty:

  • Breathing rate

  • Battery life

  • Historical data

  • Humidity

  • Mask change detection

Mid-Fi Prototype

As a UI designer, I choose blue and white as the main color to convey an idea of health. I show all the main data on the first page home page to make it more informative.

On the top bar is the name of the device, which users connect to. And I use the small point to show the connection state: green means connected, grey means unconnected, and red means low battery level.

A/B Test

We did A/B Test on How historical data is displayed:

  1. One method is for the historical data to drop down on the same page, keeping everything in a single view.

  2. The second method is for an interaction to take the user to a subpage.

Overall, we received 99 responses. We will first analyze each question as a whole, and then compare survey A with survey B. For analysis, we used t-tests and Net Promoter Score to verify our hypothesis. The A/B Test result shows that:

  1. Participants believe both version A and version B are easy to read information.

  2. Participants tend to believe that version A is more informative than version B.

High-fi Prototype

I developed the high-fi app prototype on Android studio. 

We used SQLite as a database and UART to provide a low-energy Bluetooth connection.


Hardware Architecture

Software Architecture

With-in Subject Experiments


In order to verify whether the Zelypso masks we designed can really make people feel comfortable, we recruited 10 participants with no respiratory health issues.

Participants were asked to wear different masks (or no masks) and report their feeling before and after exercise. And we counted their breath cycles and breath rates.


  1. Zelypso Mask with Fan was rated most comfortable pre- and post-exercise outside of not wearing a mask
  2. Zelypso Mask with No Fan had the biggest drop in comfort rating pre- and post-exercise
  3. Different fabrics can reduce effectiveness of fan drastically
  4. In general, breathing rates increase as more layers added

Affinity Diagraming