Interaction Design 3 — Spring 2025

KnowNow

Mobile App Emergency UX 2025

When time matters, language shouldn't be a barrier. A mobile app designed to help non-English speaking communities access emergency alerts, shelter information, and evacuation guidance in their native language.

Role
Interaction Design, User Research
Team
Lanna, Silvia, Schteffen
Duration
14 weeks, Spring 2025
Tools
Figma, Paper Prototyping
Methods
Field Research, User Testing
Course
Interaction Design 3

Emergency alerts were built for English speakers

During the 2025 LA wildfires, over 12,000 of 50,000 Asian residents in four evacuation zones needed language assistance — yet alerts arrived in English and Spanish only. For non-English speaking communities, a language barrier in an emergency isn't an inconvenience. It's a life-safety failure.

Language access plans
67%
of U.S. counties recognize language access in emergency plans
LEP households impacted
4,178
LEP households impacted in Eaton fire zone alone
English literacy gap
30M+
adults in the U.S. who cannot read or write in English
Secondary research findings — language gaps in emergency infrastructure

Secondary research findings — language gaps in emergency infrastructure

We went to where the crisis actually was

We visited an active Red Cross emergency shelter, recruited participants through social media and physical posters in the community, and reached out via text to first responders, bilingual volunteers, and emergency workers across LA. What we found shaped everything.

Field visit — Red Cross emergency shelter, Los Angeles, January 2025

Field visit — Red Cross emergency shelter, Los Angeles, January 2025

12,000 of 50,000 Asian residents in the evacuation zones needed language assistance — alerts came only in English and Spanish.

"Most of the alerts came through our phones eventually, but they were in English, and understanding them was hard for me... I ended up calling a friend from the cultural association to help translate."

— HUA LI

6,387 non-English speaking residents in the Eaton Fire struggled to verify information accuracy due to language barriers.

"I didn't get any official information right away... when I do watch TV, it's usually in Spanish because my English is still hard for me sometimes."

— MARIA LOPEZ (synthetic user)

26.6% of California's population are LEP. Most LA fire alerts were purely in English text.

"Everything seemed designed for people who already had more resources or who spoke the language perfectly."

— MARIA LOPEZ (synthetic user)

The most vulnerable user: a recent immigrant

Our user type matrix mapped four user types across two axes — English proficiency and disaster readiness. We focused on the highest-risk quadrant: limited English, no disaster training. That led us to Li Xiao.

User type matrix — English proficiency × disaster readiness

User type matrix — English proficiency × disaster readiness

Profile
42 years old, Cantonese-speaking, from Guangzhou
Arrived 4 months ago on a family reunification visa
Low-income household, small apartment in LA's Cantonese community
No prior wildfire or earthquake experience
Motivations
Finding emergency information in his language
Getting reliable help without fear
Keeping his family safe
Learning how to prepare before the next disaster

Li Xiao gets an English-only wildfire alert at work. His wife texts him in panic from their apartment — she can see smoke. He can't tell if they need to evacuate. He opens KnowNow, which translates the alert into Cantonese, shows fire zones visually, and sends his wife a simple message in Cantonese: "Start getting ready, pack important items."

Storyboard — Li Xiao receives the alert

Storyboard — Li Xiao receives the alert

Feature
Multilingual Alerts
Real-time alerts translated to your preferred language
Feature
Shelter Finder
Nearest shelters with accessibility and pet filters
Feature
Packing List
Disaster-type packing lists with swipe micro-interactions

Paper to pixels, shaped by testing

We started with a paper prototype and tested it with 4 participants — product design students and a faculty member. Three key findings from testing directly drove our revisions into a low-fidelity digital prototype.

Paper prototype — language selection, alerts, shelter map, packing list

Paper prototype — language selection, alerts, shelter map, packing list

Finding 01
Packing list interactions
3 of 4 users didn't see the edit button and tapped directly on list items instead of using swipe to delete.
Finding 02
Alert icon visibility
2 of 4 users clicked homepage news first and took time to locate the alert icon in the top corner.
Finding 03
Shelter filter
2 of 4 users tapped directly on the map instead of using the filter menu, finding it confusing.

Low-fidelity prototype

Based on these findings, we moved into a low-fidelity digital prototype with three targeted revisions.

Low-fidelity prototype screens
Revision 01
Packing list interactions
Updated to left-swipe = delete, right-swipe = complete, tap = edit. Added disaster-type filter tabs at the top.
Revision 02
Alert navigation
Moved the alert button to the center of bottom nav and scaled up the active disaster display.
Revision 03
Shelter filter
Replaced the confusing filter menu with a Sort By dropdown and set current location as the default.

What we'd still want to know

KnowNow reached paper prototype stage. The process surfaced questions we didn't get to answer — and that feels like the honest place to end.

Open Question
Real emergency use
How would people actually use this under genuine crisis conditions?
Open Question
Expert feedback
What would people with real emergency experience critique or change?
Open Question
Universal language
How can iconography and visual language transcend linguistic backgrounds entirely?