UX Researcher informing empathetic designs powered by curiosity

Smart Shade System


Alleviate Sleep Disturbances and Improve City Life



  • My primary roles: Researcher and VUI Designer

  • Project Duration: 10 weeks (September 2017-December 2017)

  • Tools: Sketch, Google Suite, LucidChart, InVision, Keynote

  • Team: Olivia Harold, Andrew Shiau, Will Wang

  • Deliverables: UI Specs, presentation


Our prompt was to consider the Smart City and human behavior change.

We designed a Smart home shade system, Lux, that works with people’s unique schedules to alleviate disturbances of city living to improve sleep quality. Lux simulates the sunrise and sunset to create an artificial Circadian Rhythm to improve sleep quality and health. 3 components allow for graceful integration into the users life:

01. Physical Shade with speaker and colored light

02. Voice Control

03. Companion App


Secondary Research

I began by reading journalistic and academic articles about the current Smart city landscape and problem space. 

Researching Smart city technology

20171004_142333 (1).jpg

I organized with an affinity diagram to find relationships betwen existing technologies, trends, and the psychological impact of city living, based on secondary research. Together with my team, we did three 3-minute "How Might We" challenge statement sprints to tie together themes of interest. This led us to narrow in on the problem of elevated stress in city environments. 


Defining a problem space: Overstimulation


Multiple studies showed that city dwellers on average are more stimulated than their rural counterparts (The Guardian, 2014; The Wall Street Journal, 2011). Overstimulation caused by light often negatively affects mood (Lambert et. al, 2015). This is correlated with higher rates of mental illness and stress. 


Defining a population: Atypical Workers

Stress is magnified in people who have atypical (not 9-5) work hours, which makes up about 1/5 of the workforce in America and Europe (The American Prospect). This population includes nurses, security workers, bartenders, and doctors. Increased stress is due to several factors, such as an offset from a natural Circadian rhythm and lower salaries. Atypical workers are prone to serious health risks caused by abnormal melatonin levels due to light exposure, including cancer (American Journal of Preventative Medicine).
Solving the overstimulation issue for this sizable population would have substantial impact. Solving the atypical pay economy was outside the scope of our project, so we decided to explore the problem of diminished sleep quality.


How might we improve city life for all types of individuals by improving sleep quality?


Primary Research and Pivot


To access the population, I took a novel approach by staging an in-city intervention. Together with my team, I designed a bus stop installation to collect attitudinal data about working atypical hours. We observed from a car parked across the street, and realized exactly how difficult it can be to change behaviors. I received only 3 responses, even though many commuters seemed to look directly at the robot installation. This suggests that our population could be hard to access in a public sphere. Any solution to solve social isolation could prove difficult. Perhaps intervention should manifest in a different part of their day?




Using information from the in-city intervention and secondary research, we used a braiding technique to generate 30 solution concepts and sketches including social networks, engineered cities, and drone technology.




By ranking team excitement and feasibility, I motivated my team to narrow down to 2 combined concepts.


Simulated Circadian Cycle

“Natural Spectrum light strips for bedroom installation. Programmable to create a simulated sunrise and daytime according to your personal work hours. Results in more satisfying sleep and increased health and energy.”

Noise-blocking Curtains

“Cheaper, more versatile, and more eco-friendly than traditional white- noise machines. Increases sleep quality by blocking noise from roommates and environmental disturbances, or to cover your own noise from activities that don’t fit into a normal schedule. Feel free to watch that movie on surround sound when you get home from work... at 6 am!”


Usability Testing and Paper Prototypes

I designed a Usability study to test our concepts with low fidelity prototypes. The goal was to gauge higher-level reactions to the product. We tested interface and voice interactions separately on three participants.


Research Questions:

01. How can we create an efficient balance between automatic functionality and freedom for the user?

02. Are people aware of the city’s effect on their sleep quality?

03. How will we balance function and feature on the product?


Our concept was novel, and one challenge of testing was deciding how much context to give the user. In a real-life situation, our user would know our product before purchasing, but we decided to only provide usability testers with packaging for context.

Additionally, a hybrid paper prototype was employed by drawing an interface and attaching it to an iPhone alarm app to quickly recreate a tactile and responsive experience.

Participants were given specific tasks: set an alarm, change transition time, change alarm sound. A semi-structured interview following the tasks collected higher-level attitudes. 


Screen Shot 2018-05-21 at 11.12.39 PM.png



Bed time and wake time consistencies do not match. Our solution should strike a balance between persistent wake-up alarms and gentle bedtime prompts. 

I usually wake up at the same time, but bedtime can vary wildly.
— User Tester #2


Due to the newness of voice-controlled devices, some people are opposed and prefer app control. Contrarily, other people desire the convenience of voice.

I’m not a huge fan of voice control.
— User Tester #3
Why would I download an app if I can just talk to the curtain?
— User Tester #1

Design Principles

01. Make all features controllable by both app and voice.

02. Design within current behaviors.

03. Allow for preferences.


Hero Journeys and Wireframes

My team had many discussions and our own assumptions about the details of the product, which often led to backtracking and confusion. To specify the automated behaviors of the physical shade and understand the define the functionality, I created wireframes and matched them to a timeline of environmental factors. It clarified details about the interaction between the app, shade, and environment to the team as well as instructors and classmates who reviewed our project. We chose three flows: Atypical, Nap, and Typical. 


01. Atyical User: Night Shift

Atypical User_ Night Shift.png

02. Occasional User: Naps

Occasional User_ Nap.png

03. Typical User: 9-5

Typical User_ 9-5.png

Low Fidelity App Flow

To show flow and functionality throughout the app to simplify development.


Voice Control

On-going research into Voice User Interfaces led me to realize that the bedroom is the ideal application for voice control:

  • Quiet environment

  • For simple tasks

  • Screenless control, lowering pre-bedtime exposure to blue light

  • Less steps than app control


Versus app control, which requires unlocking phone, opening the Lux app, navigating to the “remote” screen, and adjusting the controller for the light.

Versus app control, which requires unlocking phone, opening the Lux app, navigating to the “remote” screen, and adjusting the controller for the light.


The specifications consisted of a flow chart mapping likely scenarios. Necessary functions were based on conversations with users during testing, and looking at similar products like voice-controlled home assistants. 

VoiceTree-275 copy.png


I found myself operating as an advocate for the user. When me and my teammates disagreed
on common bedtime habits, I would suggest polling the room, or guerilla usability testing,
rather than making a decision for potential users. 

Near the end of the project, I realized that the 9-5 user flows also did not match to a natural
Circadian rhythm. The system we developed was generalizable, increasing the potential impact
more than we conceived when we began.


Future Directions

01. Utilizing Machine Learning to learn users patterns for full home automation.

02. Working prototype and longitudinal use for further improvement of usability, and measurement of health impact.