Ambient Guidance, Spatial Communication
Ambient Guidance, Spatial Communication
AI Elevator Mirror Display
that Enables Smart Home Experiences
AI Elevator Mirror Display
that Enables Smart Home Experiences
AI Elevator Mirror Display
that Enables Smart Home Experiences
In collaboration with
In collaboration with
In collaboration with
Electronics (UX Lab) through an industry-academia partnership.
Electronics (UX Lab) through an industry-academia partnership.
Electronics (UX Lab) through an industry-academia partnership.
Problem
Problem
Problem
Users don’t get to experience smart home features nor get used to them.
Users don’t get to experience smart home features nor get used to them.
Users don’t get to experience smart home features nor get used to them.
Most smart home functions reveal their benefits only after being used, which gives users little motivation to try them in the first place.
Most smart home functions reveal their benefits only after being used, which gives users little motivation to try them in the first place.
Most smart home functions reveal their benefits only after being used, which gives users little motivation to try them in the first place.
Solution
Solution
Solution
Elevator moments let them discover these features naturally and without effort.
Elevator moments let them discover these features naturally and without effort.
Elevator moments let them discover these features naturally and without effort.
Without having to search for anything, they experience these functions lightly within their daily routine,
helping them recognize their value without resistance.
Without having to search for anything, they experience these functions lightly within their daily routine, helping them recognize their value without resistance.
Without having to search for anything, they experience these functions lightly within their daily routine, helping them recognize their value without resistance.
Problem
Problem
Problem
*User Interview : 2023.07.20~24, 8 Participants
*User Interview : 2023.07.20~24, 8 Participants
Many smart home features exist, but users rarely get to learn them.
Many smart home features exist, but users rarely get to learn them.
Many smart home features exist, but users rarely get to learn them.
Literature Research
Literature Research
Literature Research
Learning happens in small repeated moments, not in study time.
Learning happens in small repeated moments, not in study time.
Learning happens in small repeated moments, not in study time.
Khan, A. A. (2025). Assessing student learning by reducing cognitive load in postgraduate fire safety engineering courses.
Discover Education, 4, Article 284. https://doi.org/10.1007/s44217-025-00724-y
Khan, A. A. (2025). Assessing student learning by reducing cognitive load in postgraduate fire safety engineering courses.
Discover Education, 4, Article 284. https://doi.org/10.1007/s44217-025-00724-y
Current User Flow
Current User Flow
Current User Flow
*User Interview : 2023.07.20~24, 8 Participants
*User Interview : 2023.07.20~24, 8 Participants
*User Interview : 2023.07.20~24, 8 Participants
Repeated content consumption during elevator waiting.
Repeated content consumption during elevator waiting.
Repeated content consumption during elevator waiting.
Short but frequent moments that support ‘learning new features’.
Short but frequent moments that support ‘learning new features’.
Short but frequent moments that support ‘learning new features’.
Survey
Survey
Survey
*Survey : 2023.07.10~07.16, 86 Participants
*Survey : 2023.07.10~07.16, 86 Participants
The elevator as a potential space for new learning experiences.
The elevator as a potential space for new learning experiences.
The elevator as a potential space for new learning experiences.
Survey reveals the elevator’s potential as a tool for better smart home use.
Survey reveals the elevator’s potential as a tool for better smart home use.
77.8%
Willing to accept feature tips if they require no extra effort.
Willing to accept feature tips if they require no extra effort.
Willing to accept feature tips if they require no extra effort.
61.2%
If I see it repeatedly, I might try the feature at least once.
If I see it repeatedly, I might try the feature at least once.
If I see it repeatedly, I might try the feature at least once.
50.8%
I look at the screen content or guidance while waiting.
I look at the screen content or guidance while waiting.
I look at the screen content or guidance while waiting.
How might we + Success Metrics
How might we + Success Metrics
How might we + Success Metrics
User Test & Physical Environment Study
User Test & Physical Environment Study
User Test & Physical Environment Study
Designing Low-Burden Interactions Through On-Site Testing.
Designing Low-Burden Interactions Through On-Site Testing.
Designing Low-Burden Interactions Through On-Site Testing.
Duration
2023.07 (1 day on-site + 2 days analysis and user testing)
Purpose
To verify what UI size, placement, and frequency feel natural in the confined elevator space through measurements, paper prototypes, and user feedback.
Duration
2023.07 (1 day on-site + 2 days analysis and user testing)
Purpose
To verify what UI size, placement, and frequency feel natural in the confined elevator space through measurements, paper prototypes, and user feedback.
Contents
Measuring the elevator interior
Creating UI-sized paper prototypes
Testing prototypes on-site based on user sightlines
Brief interviews on preferred UI size, frequency, and information load
Contents
Measuring the elevator interior
Creating UI-sized paper prototypes
Testing prototypes on-site based on user sightlines
Brief interviews on preferred UI size, frequency, and information load
*Permission to use interviewee photos in this portfolio has been obtained.
*Permission to use interviewee photos in this portfolio has been obtained.
Do
Do
Do
Offer small, glanceable bits of information
Keep AI involvement minimal and hint-like
Use natural, context-based recommendations
Show only one piece of information at a time
Maintain quiet interactions that don’t interrupt
Offer small, glanceable bits of information
Keep AI involvement minimal and hint-like
Use natural, context-based recommendations
Show only one piece of information at a time
Maintain quiet interactions that don’t interrupt
Offer small, glanceable bits of information
Keep AI involvement minimal and hint-like
Use natural, context-based recommendations
Show only one piece of information at a time
Maintain quiet interactions that don’t interrupt
Don’t
Don’t
Don’t
Use large or flashy visuals that look like ads
Show complex or effort-heavy information
Show camera-based or user-targeting UI
Use excessive pop-ups or repeated alerts
Push messages that assume user intent
Use large or flashy visuals that look like ads
Show complex or effort-heavy information
Show camera-based or user-targeting UI
Use excessive pop-ups or repeated alerts
Push messages that assume user intent
Use large or flashy visuals that look like ads
Show complex or effort-heavy information
Show camera-based or user-targeting UI
Use excessive pop-ups or repeated alerts
Push messages that assume user intent
Voice AI Persona
Voice AI Persona
Voice AI Persona
AI conversation design that reaches emotions first, without feeling intrusive.
AI conversation design that reaches emotions first, without feeling intrusive.
AI conversation design that reaches emotions first, without feeling intrusive.
Communicate with gentle and understanding language, considering even the first impressions of non–smart home users.
Communicate with gentle and understanding language, considering even the first impressions of non–smart home users.
Fewer than 3 voice prompts: Too much guidance feels tiring, but long silence feels awkward
Fewer than 3 voice prompts: Too much guidance feels tiring, but long silence feels awkward
Fewer than 3 voice prompts: Too much guidance feels tiring, but long silence feels awkward
Stages
Stages
Stages
1
1
1
2
2
2
3
3
3
Trigger design points
Trigger design points
Trigger design points
Right after the door opens
Right after the door opens
Right after the door opens
During movement
During movement
During movement
Right before arrival
Right before arrival
Right before arrival
When two or more people are onboard
When two or more people are onboard
When two or more people are onboard
4
4
4
Within 1–2 seconds after boarding
Within 1–2 seconds after boarding
Within 1–2 seconds after boarding
One prompt during quiet moments
One prompt during quiet moments
One prompt during quiet moments
Two seconds before arrival
Two seconds before arrival
Two seconds before arrival
Deactivated / kept neutral when needed
Deactivated / kept neutral when needed
Deactivated / kept neutral when needed
Speech timing
Speech timing
Speech timing
Welcome!
Welcome!
Welcome!
Should I start your car?
Should I start your car?
Should I start your car?
I’ve turned on some music.
I’ve turned on some music.
I’ve turned on some music.
(No speech)
(No speech)
(No speech)
Examples
Examples
Examples
Avoid explicit personal data and direct statements.
Avoid explicit personal data and direct statements.
Avoid explicit personal data and direct statements.
✅ Information that may be spoken
✅ Information that may be spoken
✅ Information that may be spoken
SmartThings-related interactions
Real-time transit updates / remote car start
Floor arrival guidance
Music / movie recommendations
Short, calming messages
SmartThings-related interactions
Real-time transit updates / remote car start
Floor arrival guidance
Music / movie recommendations
Short, calming messages
SmartThings-related interactions
Real-time transit updates / remote car start
Floor arrival guidance
Music / movie recommendations
Short, calming messages
Full name + unit number
Direct emotional statements (“You look depressed.”)
Personal information when two or more people are onboard
Full name + unit number
Direct emotional statements (“You look depressed.”)
Personal information when two or more people are onboard
Full name + unit number
Direct emotional statements (“You look depressed.”)
Personal information when two or more people are onboard
❌ Information that should not be spoken
❌ Information that should not be spoken
❌ Information that should not be spoken
Speech design : allows simple emotional touchpoints.
Speech design : allows simple emotional touchpoints.
Speech design : allows simple emotional touchpoints.
Serious
Serious
Serious
Fast
Fast
Fast
Casual
Casual
Casual
Friendly
Friendly
Friendly
Descriptive
Descriptive
Descriptive
Cheerful
Cheerful
Cheerful
Slow
Slow
Slow
Formal
Formal
Formal
Polite
Polite
Polite
Intuitive
Intuitive
Intuitive
Specialized Functions
Specialized Functions
Specialized Functions
Personalized guidance based on user characteristics
Personalized guidance based on user characteristics
Personalized guidance based on user characteristics
Smart home appliance control
Smart home appliance control
Smart home appliance control
Goals
Goals
Goals
Quickly understand user needs and provide attentive guidance
Ensure users feel comfortable when interacting with the system
Quickly understand user needs and provide attentive guidance
Ensure users feel comfortable when interacting with the system
Quickly understand user needs and provide attentive guidance
Ensure users feel comfortable when interacting with the system
User Interaction Scenario
User Interaction Scenario
User Interaction Scenario
Simple smart home experiences that blend into everyday life.
Simple smart home experiences that blend into everyday life.
Simple smart home experiences that blend into everyday life.
Service Modeling
Service Modeling
Service Modeling
From first use to reuse : A journey that makes smart home use easier.
From first use to reuse : A journey that makes smart home use easier.
From first use to reuse : A journey that makes smart home use easier.
When going out in the morning
When going out in the morning
When going out in the morning
When coming home in the evening
When coming home in the evening
When coming home in the evening
Impact & Result
Impact & Result
Impact & Result
Adoption: Increased willingness to use smart home features
Adoption: Increased willingness to use smart home features
Adoption: Increased willingness to use smart home features
133% Improvement in intention.
133% Improvement in intention.
133% Improvement in intention.
Q. Are you willing to use smart home features?
Q. Are you willing to use smart home features?
Q. Are you willing to use smart home features?
20 participants
20 participants
20 participants
Before
Before
Before
1.8/5
1.8/5
1.8/5
After
After
After
4.2/5
4.2/5
4.2/5
I wasn’t interested in smart home features before, but seeing them presented so naturally makes me want to try them.
I wasn’t interested in smart home features before, but seeing them presented so naturally makes me want to try them.
I wasn’t interested in smart home features before, but seeing them presented so naturally makes me want to try them.
I didn’t even know this feature existed. I usually avoid things that feel like work, so it was nice to learn about it in such an effortless way.
I didn’t even know this feature existed. I usually avoid things that feel like work, so it was nice to learn about it in such an effortless way.
I didn’t even know this feature existed. I usually avoid things that feel like work, so it was nice to learn about it in such an effortless way.
Accessibility
Accessibility
Accessibility
Accessibility as a way to amplify user strengths.
Accessibility as a way to amplify user strengths.
Accessibility as a way to amplify user strengths.
Mentor Feedback
Mentor Feedback
Mentor Feedback
Expanding the team’s perspective beyond existing frames.
Expanding the team’s perspective beyond existing frames.
Expanding the team’s perspective beyond existing frames.
Yunjin widened our team’s perspective by offering insights that were uncommon even inside LG. Her idea of turning brief everyday moments into smart-home touchpoints was new to existing research and opened a fresh direction for our project. Throughout the work, she showed a clear ability to create original concepts from familiar problems, and it left a strong impression on us.
Yunjin widened our team’s perspective by offering insights that were uncommon even inside LG. Her idea of turning brief everyday moments into smart-home touchpoints was new to existing research and opened a fresh direction for our project. Throughout the work, she showed a clear ability to create original concepts from familiar problems, and it left a strong impression on us.
Yunjin widened our team’s perspective by offering insights that were uncommon even inside LG. Her idea of turning brief everyday moments into smart-home touchpoints was new to existing research and opened a fresh direction for our project. Throughout the work, she showed a clear ability to create original concepts from familiar problems, and it left a strong impression on us.
* Mentor. Kim : Senior UX Designer, LG Electronics
* Mentor. Kim : Senior UX Designer, LG Electronics
What I’ve Learned
What I’ve Learned
What I’ve Learned
When users don’t have to decide to use it, technology feels easier to accept.
When users don’t have to decide to use it, technology feels easier to accept.
When users don’t have to decide to use it, technology feels easier to accept.
Smart home features require users to open an app and search for functions on their own, which creates small but constant effort. But the elevator is a moment where people normally do nothing, and having information there does not feel strange. As a result, users can come across the functions without any intention to use them, and this creates almost no resistance. Through this experience, I learned that a low-resistance UX comes from designing touchpoints where users meet information naturally, without needing to participate consciously.
Smart home features require users to open an app and search for functions on their own, which creates small but constant effort. But the elevator is a moment where people normally do nothing, and having information there does not feel strange. As a result, users can come across the functions without any intention to use them, and this creates almost no resistance. Through this experience, I learned that a low-resistance UX comes from designing touchpoints where users meet information naturally, without needing to participate consciously.
Smart home features require users to open an app and search for functions on their own, which creates small but constant effort. But the elevator is a moment where people normally do nothing, and having information there does not feel strange. As a result, users can come across the functions without any intention to use them, and this creates almost no resistance. Through this experience, I learned that a low-resistance UX comes from designing touchpoints where users meet information naturally, without needing to participate consciously.
More projects
Hardware Interaction
Hardware Interaction
A baby's cry causes analysis using a wearable device and an application
A baby's cry causes analysis using a wearable device and an application
A baby's cry causes analysis using a wearable device and an application
Helping understand babies better
Helping understand babies better
Helping understand babies better
Accessibility
Accessibility
Subtitles that let hearing-impaired viewers immerse in the movie’s emotions & context
Subtitles that let hearing-impaired viewers immerse in the movie’s emotions & context
Subtitles that let hearing-impaired viewers immerse in the movie’s emotions & context
User satisfaction improved by 153%
User satisfaction improved by 153%
Satisfaction improved by 153%
Sponsored project, Hardware
Sponsored project, Hardware
Ideation Meeting Support Tool for ADHD
Ideation Meeting Support Tool for ADHD
Ideation Meeting Support Tool for ADHD
Helping Users 59.3% More Focused
Helping Users 59.3% More Focused
Helping Users 59.3% More Focused
