Designing proactive AI support
for in-flight motion sickness
Designing proactive AI support
for in-flight motion sickness
Designing proactive AI support for in-flight
motion sickness
UI/UX Design
UI/UX Design
Hardware-Software Interaction
Hardware-Software Interaction
Wearable Interaction Strategy
Wearable Interaction Strategy
Role:
Product Designer (Team Lead)
Team:
5-Person Interdisciplinary Team
Timeline:
5 Weeks // 2025
Context:
UW MHCI+D Project
Role:
Product Designer (Team Lead)
Team:
5-Person Interdisciplinary Team
Timeline:
5 Weeks // 2025
Context:
UW MHCI+D Project
Navigation
Imagine spending an entire flight waiting to feel sick.
Imagine spending an entire flight waiting to feel sick.
Turbulence, visual motion, fatigue, dehydration, and anxiety can stack together in the air. Once discomfort begins, passengers cannot stop the motion, leave the environment, or easily regain control.
Turbulence, visual motion, fatigue, dehydration, and anxiety can stack together in the air. Once discomfort begins, passengers cannot stop the motion, leave the environment, or easily regain control.
The design gap:
The design gap:
Current remedies are fragmented and reactive: leaving passengers to predict symptoms, coordinate support, and manage discomfort on their own.
Current remedies are fragmented and reactive: leaving passengers to predict symptoms, coordinate support, and manage discomfort on their own.
// The Design
// The Design
Aura is a speculative connected system designed
to identify rising motion-sickness risk and support passengers before symptoms escalate.
Aura is a speculative connected system designed
to identify rising motion-sickness risk and support passengers before symptoms escalate.
Aura is a speculative connected system designed to identify rising motion-sickness risk and support passengers before symptoms escalate.
The earbuds monitor physical and flight conditions, the app gives passengers control, and the AI connects those inputs to recommend timely support. Passengers can review, adjust, or dismiss each recommendation.
The earbuds monitor physical and flight conditions, the app gives passengers control, and the AI connects those inputs to recommend timely support. Passengers can review, adjust, or dismiss each recommendation.
Aura App
Aura App
Personalize support, review recommendations, and control the connected system.
Personalize support, review recommendations, and control the connected system.
AuraBuds
AuraBuds
Monitor physical and flight conditions and deliver on-demand sound therapy.
Monitor physical and flight conditions and deliver on-demand sound therapy.
Aura AI
Aura AI
Connect passenger and flight context to identify rising risk and explain what to do next.
Connect passenger and flight context to identify rising risk and explain what to do next.
// Market Signal
// Market Signal
Six months later, Samsung entered the same
motion-sickness product space.
Six months later, Samsung entered the same
motion-sickness product space.
Samsung released Hearapy, an app that delivers a 100 Hz tone through compatible earbuds to help reduce motion sickness. The product draws from the same Nagoya University research that informed Aura.
Samsung released Hearapy, an app that delivers a 100 Hz tone through compatible earbuds to help reduce motion sickness. The product draws from the same Nagoya University research that informed Aura.
Article from 9to5Google
Article from 9to5Google
Hearapy provided an external market signal for earbud-based motion-sickness support. While Samsung focused on a user-initiated audio session, Aura explored how sensing, timing, and user-controlled recommendations could work together as one connected system.
Hearapy provided an external market signal for earbud-based motion-sickness support. While Samsung focused on a user-initiated audio session, Aura explored how sensing, timing, and user-controlled recommendations could work together as one connected system.
// My Contributions
// My Contributions
As team lead, I helped narrow a broad exploration of travel discomfort into a focused question: how could a connected system support passengers before motion sickness escalated?
Wrote the PRD and aligned the app, hardware, and AI workstreams.
Led research synthesis and translated insights into four design criteria.
Refined the app architecture, defined the earbud hardware requirements, and designed the system’s support logic.
As team lead, I helped narrow a broad exploration of travel discomfort into a focused question: how could a connected system support passengers before motion sickness escalated?
Wrote the PRD and aligned the app, hardware, and AI workstreams.
Led research synthesis and translated insights into four design criteria.
Refined the app architecture, defined the earbud hardware requirements, and designed the system’s support logic.
My highest-leverage contribution was shaping three decisions that defined the final concept:
My highest-leverage contribution was shaping three decisions that defined the final concept:
// Decision 1
From standalone remedies to one connected system
I defined four design criteria that helped the team converge on a coordinated hardware–software experience rather than a collection of isolated features.
// Decision 1
From standalone remedies to one connected system
I defined four design criteria that helped the team converge on a coordinated hardware–software experience rather than a collection of isolated features.
// Decision 1
From standalone remedies to one connected system
I defined four design criteria that helped the team converge on a coordinated hardware–software experience rather than a collection of isolated features.
// Decision 2
From generic earbuds to purpose-built hardware
I translated unmet in-flight needs into sensing and relief requirements that an app or conventional earbuds could not address alone.
// Decision 2
From generic earbuds to purpose-built hardware
I translated unmet in-flight needs into sensing and relief requirements that an app or conventional earbuds could not address alone.
// Decision 2
From generic earbuds to purpose-built hardware
I translated unmet in-flight needs into sensing and relief requirements that an app or conventional earbuds could not address alone.
// Decision 3
From reactive controls to explainable support logic
I designed a deterministic model that combined context, timing, and user state to recommend support while keeping every action reviewable and dismissible.
// Decision 3
From reactive controls to explainable support logic
I designed a deterministic model that combined context, timing, and user state to recommend support while keeping every action reviewable and dismissible.
// Decision 3
From reactive controls to explainable support logic
I designed a deterministic model that combined context, timing, and user state to recommend support while keeping every action reviewable and dismissible.
// Decision 1
// Decision 1
Why a connected system over a simpler standalone solution?
Why a connected system over a simpler standalone solution?
User interview + Understand motion sickness + Competitive analysis
The research pointed to four experience requirements: support had to arrive before symptoms escalated, require little screen use, remain discreet in a shared cabin, and adapt to changing flight and passenger conditions.
The research pointed to four experience requirements: support had to arrive before symptoms escalated, require little screen use, remain discreet in a shared cabin, and adapt to changing flight and passenger conditions.
The research pointed to four experience requirements: support had to arrive before symptoms escalated, require little screen use, remain discreet in a shared cabin, and adapt to changing flight and passenger conditions.
I translated those insights into four design criteria and used them to evaluate every concept across the app, hardware, and AI workstreams.
I translated those insights into four design criteria and used them to evaluate every concept across the app, hardware, and AI workstreams.
Team ideation synthesis
Team ideation synthesis
Team ideation synthesis
After synthesizing our ideation outcomes, the team aligned on the connected system because it has the most completed solution that can satisfy all 4 design criteria, despite being more challenging to design and build.
After synthesizing our ideation outcomes, the team aligned on the connected system because it has the most completed solution that can satisfy all 4 design criteria, despite being more challenging to design and build.
Direction
Pros
Cons
Software-only solutions with passenger's own earbuds
Software-only solutions with passenger's own earbuds
Easy to access and control, no extra hardware
Easy to access and control, no extra hardware
Too manual; the screen itself can worsen nausea
Too manual; the screen itself can worsen nausea
Hardware-only solutions
Hardware-only solutions
No screens, focused and direct relief
No screens, focused and direct relief
Single-purpose; one more item to carry
Single-purpose; one more item to carry
Hardware and software connected solution
Hardware and software connected solution
Enable systemic sensing and relieving
Enable systemic sensing and relieving
More complex to build and design
More complex to build and design
// Decision 2
// Decision 2
Why custom-made earbuds are needed?
Why custom-made earbuds are needed?
Research paper on 100 Hz and motion sickness
During ideation, a teammate uncovered research from research from Nagoya University Graduate School of Medicine suggesting that a short 100 Hz audio intervention might reduce motion-sickness symptoms. But because any compatible earbuds could deliver the tone, I questioned whether Aura should introduce another physical product at all.
During ideation, a teammate uncovered research from research from Nagoya University Graduate School of Medicine suggesting that a short 100 Hz audio intervention might reduce motion-sickness symptoms. But because any compatible earbuds could deliver the tone, I questioned whether Aura should introduce another physical product at all.
My industrial design background led me to look beyond audio output. I explored whether purpose-built earbuds could address additional in-flight discomfort, collect relevant context, and let passengers respond without returning to a screen.
My industrial design background led me to look beyond audio output. I explored whether purpose-built earbuds could address additional in-flight discomfort, collect relevant context, and let passengers respond without returning to a screen.
Adapting a pressure-regulating filter for AuraBuds
Adapting a pressure-regulating filter for AuraBuds
In our competitive research, EarPlanes stood out because its micro-ceramic filter is designed to slow barometric pressure changes, helping passengers reduce ear pain and pressure discomfort during flight.
In our competitive research, EarPlanes stood out because its micro-ceramic filter is designed to slow barometric pressure changes, helping passengers reduce ear pain and pressure discomfort during flight.
Passengers also described ear pressure as an overlapping source of in-flight discomfort. Although separate from motion sickness itself, reducing that additional burden could make AuraBuds more useful during the moments when passengers already feel vulnerable.
Passengers also described ear pressure as an overlapping source of in-flight discomfort. Although separate from motion sickness itself, reducing that additional burden could make AuraBuds more useful during the moments when passengers already feel vulnerable.
Testing the EarPlane ceramic filter
Testing the EarPlane ceramic filter
I used the pressure change in the Northgate–Roosevelt light-rail tunnel as a repeatable experience probe. It did not validate clinical effectiveness, but it let me experience the filter mechanism directly and explore how pressure regulation could become part of the earbud architecture.
I used the pressure change in the Northgate–Roosevelt light-rail tunnel as a repeatable experience probe. It did not validate clinical effectiveness, but it let me experience the filter mechanism directly and explore how pressure regulation could become part of the earbud architecture.
I used the pressure change in the Northgate–Roosevelt light-rail tunnel as a repeatable experience probe. It did not validate clinical effectiveness, but it let me experience the filter mechanism directly and explore how pressure regulation could become part of the earbud architecture.
This shifted AuraBuds from a proprietary audio output into a flight-specific interaction surface combining pressure management, sensing, and low-attention control.
This shifted AuraBuds from a proprietary audio output into a flight-specific interaction surface combining pressure management, sensing, and low-attention control.
Audio intervention
Audio intervention
Pressure management
Pressure management
Context and control
Context and control
AuraBuds
AuraBuds
Can prepare and deliver the intervention through system recommendations
Can prepare and deliver the intervention through system recommendations
Proposed integrated pressure-regulating filter
Proposed integrated pressure-regulating filter
Proposed flight-specific sensing and low-attention interaction
Proposed flight-specific sensing and low-attention interaction
Standard earbuds
Can deliver the same 100 Hz tone
Can deliver the same 100 Hz tone
Can deliver the same 100 Hz tone
Not typically designed for pressure regulation
Not typically designed for pressure regulation
General-purpose sensors and controls
General-purpose sensors and controls
// Decision 3
// Decision 3
How is the AI logic system designed?
Usability testing and concept walkthrough
We tested a mid-fidelity prototype and concept walkthrough with 8 participants to evaluate whether they understood the connected system and knew when and how to act on its recommendations.
We tested a mid-fidelity prototype and concept walkthrough with 8 participants to evaluate whether they understood the connected system and knew when and how to act on its recommendations.
6 of 8 participants expected Aura to recognize changing conditions and offer support without requiring them to search for a remedy.
6 of 8 participants expected Aura to recognize changing conditions and offer support without requiring them to search for a remedy.
AI logic system ideation
AI logic system ideation
Information architecture restructure
Information architecture restructure
I designed rule-based recommendation logic that defines when Aura should offer support, what it should explain, and which decisions remain with the passenger.
I designed rule-based recommendation logic that defines when Aura should offer support, what it should explain, and which decisions remain with the passenger.
1 // Sense
Collect contextual and biometric signals from the flight environment with AuraBuds' sensors.
1 // Sense
Collect contextual and biometric signals from the flight environment with AuraBuds' sensors.
1 // Sense
Collect contextual and biometric signals from the flight environment with AuraBuds' sensors.
2 // Interpret
Compare those signals against flight conditions, user history, and personal preferences.
2 // Interpret
Compare those signals against flight conditions, user history, and personal preferences.
2 // Interpret
Compare those signals against flight conditions, user history, and personal preferences.
3 // Recommend
Suggest relevant support such as sound therapy, breathing guidance, or hydration reminders.
3 // Recommend
Suggest relevant support such as sound therapy, breathing guidance, or hydration reminders.
3 // Recommend
Suggest relevant support such as sound therapy, breathing guidance, or hydration reminders.
4 // Explain
Tell passengers what Aura detected, what data was used, and why a recommendation was made.
4 // Explain
Tell passengers what Aura detected, what data was used, and why a recommendation was made.
4 // Explain
Tell passengers what Aura detected, what data was used, and why a recommendation was made.
The idea: AuraBuds would collect contextual and biometric signals during flight, while AI interpreted them alongside the user’s preferences, history, and flight context inside the Aura App.
Based on this, Aura would recommend a remedy and explain why. The user could accept, choose an alternative, or dismiss it, and each choice would help refine future recommendations. This made Aura proactive without removing user control.
The idea: AuraBuds would collect contextual and biometric signals during flight, while AI interpreted them alongside the user’s preferences, history, and flight context inside the Aura App.
Based on this, Aura would recommend a remedy and explain why. The user could accept, choose an alternative, or dismiss it, and each choice would help refine future recommendations. This made Aura proactive without removing user control.
This AI logic system lets Aura offer discreet, transparent support at the right moment, while keeping the passenger in control.
This AI logic system lets Aura offer discreet, transparent support at the right moment, while keeping the passenger in control.
Finalized context-to-remedy decision condition
Finalized context-to-remedy decision condition
Finalized AI logic system
Finalized AI logic system
App UI refinement
// Final Design
// Final Design
Aura: One connected system, supporting passengers across the flight journey.
Aura: One connected system, supporting passengers across the flight journey.
Aura: One connected system, supporting passengers across the flight journey.
Context-aware · Proactive · Low effort · Discreet
Context-aware · Proactive · Low effort · Discreet
Pre-Flight Support
Pre-Flight Support
Pre-Flight Support
Pre-Flight Support
Before boarding, Aura analyzes flight details, seat location, and the passenger's motion-sickness risk profile.
For example, once the passenger enters their flight information into the Aura App, if it identifies a rear seat, Aura suggests a better option, since research shows rear seats increase the likelihood of motion sickness.
This is the support Aura provides before the passenger ever enters the cabin.
Before boarding, Aura analyzes flight details, seat location, and the passenger's motion-sickness risk profile.
For example, once the passenger enters their flight information into the Aura App, if it identifies a rear seat, Aura suggests a better option, since research shows rear seats increase the likelihood of motion sickness.
This is the support Aura provides before the passenger ever enters the cabin.
Takeoff Support
Takeoff Support
Takeoff Support
Takeoff Support
During takeoff, Aura combines flight-stage awareness with body signals to offer support when stress, noise, and pressure changes are most likely to occur.
AuraBuds' active noise canceling reduces engine noise, the ceramic filter eases pressure transitions, and built-in sound therapy delivers calming, guided relief.
During takeoff, Aura combines flight-stage awareness with body signals to offer support when stress, noise, and pressure changes are most likely to occur.
AuraBuds' active noise canceling reduces engine noise, the ceramic filter eases pressure transitions, and built-in sound therapy delivers calming, guided relief.
In-Flight Support
In-Flight Support
In-Flight Support
In-Flight Support
Mid-flight, Aura uses live turbulence data and the passenger's profile to prepare them before symptoms build.
A quiet earbud reminder flags upcoming turbulence, and a simple double tap lets the passenger activate 100Hz sound therapy or their own remedy, turning an unpredictable flight into something more manageable.
Mid-flight, Aura uses live turbulence data and the passenger's profile to prepare them before symptoms build.
A quiet earbud reminder flags upcoming turbulence, and a simple double tap lets the passenger activate 100Hz sound therapy or their own remedy, turning an unpredictable flight into something more manageable.
// Reflection
// Reflection
Five weeks is enough time to build a credible concept, but not enough to validate one.
Five weeks is enough time to build a credible concept, but not enough to validate one.
There are more work required if Aura is in a real product development process:
There are more work required if Aura is in a real product development process:
Biometric Validation
The AI sensing model relies on heart rate, HRV, and movement data, but validating whether these signals can predict motion sickness (at what thresholds) would require real flight testing with motion-sensitive participants.
Biometric Validation
The AI sensing model relies on heart rate, HRV, and movement data, but validating whether these signals can predict motion sickness (at what thresholds) would require real flight testing with motion-sensitive participants.
Biometric Validation
The AI sensing model relies on heart rate, HRV, and movement data, but validating whether these signals can predict motion sickness (at what thresholds) would require real flight testing with motion-sensitive participants.
AI Timing Validation
The key challenge for Aura AI is timing: recommending support early enough to prevent escalation, but not so early that users dismiss it. This would require iterative testing across real flights.
AI Timing Validation
The key challenge for Aura AI is timing: recommending support early enough to prevent escalation, but not so early that users dismiss it. This would require iterative testing across real flights.
AI Timing Validation
The key challenge for Aura AI is timing: recommending support early enough to prevent escalation, but not so early that users dismiss it. This would require iterative testing across real flights.
Industrial Design of AuraBuds
The 3D-printed prototype validated ergonomic fit, but a production-ready earbud with a ceramic filter, biometric sensors, and a speaker driver would require further engineering feasibility work.
Industrial Design of AuraBuds
The 3D-printed prototype validated ergonomic fit, but a production-ready earbud with a ceramic filter, biometric sensors, and a speaker driver would require further engineering feasibility work.
Industrial Design of AuraBuds
The 3D-printed prototype validated ergonomic fit, but a production-ready earbud with a ceramic filter, biometric sensors, and a speaker driver would require further engineering feasibility work.
Real Usability Data at Scale
Our testing with 8 participants provided early directional feedback, but a larger research program would be needed to understand whether the AI model generalizes across real motion-sickness contexts.
Real Usability Data at Scale
Our testing with 8 participants provided early directional feedback, but a larger research program would be needed to understand whether the AI model generalizes across real motion-sickness contexts.
Real Usability Data at Scale
Our testing with 8 participants provided early directional feedback, but a larger research program would be needed to understand whether the AI model generalizes across real motion-sickness contexts.
I am still grateful to my teammates and this project because it gave me the opportunity to practice the kind of work I most want to grow in: connecting hardware, software, and AI into one coherent user experience, under constraint, with a team working across different disciplines simultaneously.
If given more time, I will explore and test more on the concept because I believe this connected system approach is promising.
I am still grateful to my teammates and this project because it gave me the opportunity to practice the kind of work I most want to grow in: connecting hardware, software, and AI into one coherent user experience, under constraint, with a team working across different disciplines simultaneously.
If given more time, I will explore and test more on the concept because I believe this connected system approach is promising.
All work © Yu-Cheng Yang // 2026