Natural Light Optimization
When a homeowner adds Lutron shades to their home, the Lutron app guides them through a quick and easy setup process that results in automatic adjustments of the shades throughout the day, letting in the right amount of light at the right time.
Core Team
UX lead, 4 product managers, 2 engineering leads, 10+ engineers
Responsibilities
Research, wireframing, user testing, motion design, visual design
Homeowners often had to adjust their Lutron shades multiple times a day because there wasn't enough natural light coming in, or there was too much natural light, making it uncomfortably bright.
How might we help homeowners automate their shades to be helpful and useful for their home needs, while maintaining a quick and simple programming experience?
“[Lutron's] Smart Wood Blinds are among the most enjoyable smart home devices I’ve used in years."
- Dan Seifert / The Verge
Picked up by YouTube reviewers and top tech blogs
Resulted in +50% increase of direct to consumer sales, 75% of total shade sales
Overview of the setup flow
Live compass
The compass gives a live reading as the user rotates the phone to get a quick and easy reading.
Map for improved accuracy
The map helps the user align themselves to their house and the window they’re facing for a more accurate reading. Rotate, drag, and zoom gestures are supported.
Start and end time
The final step is setting the timeframe the user wants the shades to actively open.
Automated schedule
After setup, a new schedule appears that adjusts the shades throughout the day as needed.
Using quantitative and qualitative research, I sought to understand the user’s pain points and synthesized them to guide us to the right solution.
Main insights
#1
90% of homeowners only have simple automation for their shades–open in the morning, close in the evening.
#2
Homeowners are annoyed about having to adjust shades throughout the day because it was too bright or too dark.
#3
Homeowners want to do more with their shades, but they don’t really know how to. It seems too complicated and tedious.
An algorithm to handle the natural light optimization
Focusing on Insight #2, engineering developed an algorithm to optimize how much natural daylight is let in throughout the day. They needed the location of the home, the time of day, position of the sun, and the direction of the shade(s).
Design challenge
Design goals
Easy to use
Accurate
Easy to understand
"The phone has a compass! Can we use that?"
Constraint: accuracy had to be within 22.5º (half a cardinal direction)
Result
The measurements are accurate 70% of the time and within the 22.5º requirement. However, the accuracy worsens depending on location, type of building, and objects nearby, with an error measurement of up to 90º.
Conclusion
We discussed the results with the stakeholders and consensed that we needed to brainstorm other ways to get the measurement.
Workshop with a cross-functional group to ideate ways to program shades to adjust with sunlight exposure, considering variables such as location, season, facade, and time of day
Sketching out top ideas narrowed down from the ideation session
Design experiments
Our top idea was to use the compass + map. However, it involved a lot of assumptions about user behavior. We wanted to do some validation before we developed this idea any further, so we quickly ran the experiment with a small pool of people in the office.
Observe how people use and orient themselves using Google Maps
<15% of people were able to identify the direction they were facing without having to use a compass or map, but used landmarks to calibrate themselves
50% of people rotated the map image to match the direction they were looking
90% of people completed the task within 10 seconds
We validated majority of our assumptions, so we continued on with this idea.
Early prototyping with more user testing
I worked with engineering to develop a preliminary version of the app to test. Users were instructed to find the direction of multiple windows in their home, and send me screenshots and a screen and sound recording of them thinking aloud as they complete the task.
Users referenced the map to confirm direction (evidenced by their think aloud)
Users said that they felt less confident in the compass reading when the map wasn’t there
All users completed the task in <15 seconds
