What if healthcare was centered around kids instead of adults?

Resmed Lumo

Research and UX Design

ResMed Respironics sponsored Carnegie Mellon University for a mobile service transformation project.  Here’s our concept Lumo. A non-intrusive pre-diagnosis system for kids and their family delivered via a mobile app. Although the sleep solution market for adults is saturated, 8-10% of children today who suffer from sleep apnea remains mostly undiagnosed.




2 Months

My Team

Calir Sun - UX & Research
Fengyi Zhang - Research
JY Park - Research

My Role

Project Lead
UX Design & Research

Problem Space

For many sleep is difficult...

Diagnosing sleep apnea in children is an akward and expensive process. Sleep Apnea is a sleep disorder in which breathing repeatedly stops and starts during sleep. Approximately 10% of children and 26% of adults suffer from Sleep Apnea. However, 90% will remain undiagnosed.

A child getting tested in a sleep lab.

And helping people sleep with technology is also difficult and tricky.

Traditionally, sleep labs are used to diagnose sleep apnea, but very few people want to undergo the experience. Today, many low-cost early detection methods come in the form of home tests and mobile apps, but both have been proven to disturb sleep regularity.

Blue light disturbs circadian rhythm.

Background Information:

What is Sleep Apnea?

It’s when your body is not getting enough oxygen while it sleeps. This usually occurs because your airway is blocked. It sounds like loud snoring.

Who is ResMed?

ResMed Inc. is a company that makes medical equipment for people struggling to breathe while sleeping. They’ve been doing this since 1989.

How Is Sleep Apnea Treated?

Continous Positive Air Pressure (CPAP) Machines are used to prevent your airway from collasping on itself.

Solution Overview

Introducing Lumo,
a non-intrusive sleep solution

Market Oppurtunity

Few have tried to design for the children’s market

After customer discovery, our team decided to focus on the children’s Market. We found that sleep apnea affects 8% to 10% of children. Moreover, many conditions worsen as these children grow up undiagnosed. And yet there are not many effective interventions to address this problem. Our solution Lumo is a family-centred solution that helps parents track children’s sleep to prediagnose potential Sleep Apnea.

The (Modified) Continuum of Health – Mental Model by Phillips Health

How It Works

Remove the screen for healthier sleep, and leverage existing sensors

By leveraging existing microphones in our smart hom we can help families and especially kids create a healthier sleep habbit.

The system is designed to work with existing smart home speakers.

Key Feature 1

Setup with any of your smart speakers

Works in the background with your smart speakers. If you do not own a smart speaker, your phone works as well.

Key Feature 2

Your family’s sleep doctor, every night

Know when to get help. Lumo works in the background to warns you of any potential sleep abnormalities in your family.

Stakeholder mapping with a vague user at the center. Hmmm...

Research and Process

Key Insight

What if instead of adults, we assume that the patient is a kid?

Through secondary research and conversation with ResMed, we found that one of the main goals is to raise people's awareness of Sleep Apnea and to diagnose more patients. To first understand ResMed structure, we first drew out a stakeholder map to define the value interchanging between them.

Turns out, it’s a huge market!

It’s 38 times amount of people in New York.


Of adults suffer from sleep apnea


Of people with obesity suffers from sleep apnea

1 in 10

Of children suffer from sleep apnea


Of people with sleep apnea remins undiagnosed

So how did we arrive at this insight?


Making + Generative Methods

Although we did a lot of exercise to uncover our assumptions and biases, it was never obvious that we assume the user was an adult until we started making. The excersise wasn't useless, because doing these generative methods helped primed our team's culture for checking our bias.

50 assumptions about healthcare and its reversal.

Questioning the idea that patients need sleep doctors to determine their sleep problem.

Me doing a quick sequence model on the whiteboard before storyboarding

Solution Overview

Concept Overview


Lower diagnosis barrier by using phone camera to measure neck size.


“Why not just just a tape measurer? It’s less akward?”


After a quiz, guide user to a certified sleep sleep lab.


“I like the certified part, but why not just use Google Maps?”


Ambient diagnosis experience for children via an IoT lamp.


“I like it, but what if I don’t want to buy a lamp?”

Concept #1

Diagnosing with Neck Measurement

This is the lowest hangest fruit. Neck size is highly correleated with sleep apnea. Turns out the ergonomics of doing this with a phone camera is just bad, plus you can just use a tape measuerer.

We roleplayed the wireframe and honestly the ergonomics of phone neck scanning were horrible.

Concept #2

Enhancing the Hands-Off Experience

After creating a sequence model and a wireframe, we realized that this domain of locating "certified sleep labs" is already captured by Google Maps. We were stuck.

Then we noticed something about the main characters of our storyboards...

We roleplayed the situation with our wireframes and Google seems own this territory.

Wait, why did we always assume “John” and “Mary” were adults and not kids?

Lesson Learned

It was through making that we made our bias visual and therefore correctible

This realization happened because by storyboarding our design solutions, we began to more clearly see how our main characters were always adults. By storyboarding, we made our bias visual and explicit and therefore correctible.

Simplified Say, Do, Make Model from Design Ethnography by me

Concept #3

Ambient Diagnosis for Kids

We didn't have time for a full storyboard, so we thumbnail the idea and role-played for quick evaluation. Considering all member's in the family, we realized that kids made up a big part of the family and realized that our whole stakeholder diagram was focused on adults.

Thumbnail Sketch of IoT Lamp

Iterative Design

From wireframes to designing high-fidelity micro-interactions

To ensure the right flow, I began with grayscale wireframes.  After the main flow is established, I handed off my wireframes to Claire for mid-fi interfaces. We collaborated on the color palette to ensure accesibility. And finally, I took charge of the hi-fi interfaces and animated our prototype with Principle.

Low Fidelity

Medium Fidelity

High Fidelity


A novel approach for a timeless company

“Although we’ve designed for families before, centering the whole experience around kids was something we never tried before but would love to!”

– Client Feedback

Thank you for reading!

Here's what I learned:

  • 🔎 Check for bias in your Design in every step of the journey
  • ✍️ How to mentor and help people grow in their desired direction
  • 🎭 Roleplaying & storyboarding is quite the power combo

Next Case Study: iRobot Internship