KnowNow

Closing the gap between alert and action

A phone lying on orange tile shows the KnowNow map screen, with the user’s location, a nearby threat, and a “You are safe” status.
overview

A multilingual, map-based emergency response app that helps people understand how a nearby threat affects them, find a safe destination, and take the next step quickly.

my contribution

KnowNow began as a 14-week team project that ended as a low-fi prototype. I then independently redesigned it, reframing the problem and rebuilding the product strategy, information architecture, interactions, and prototype.

Role
Primary Research
Product Strategy
Information Architecture
Interaction Design
Prototyping
Duration
14 weeks + independent redesign
Course
Interaction Design 3
At a glance
The interface changes as the risk changes, from monitoring to evacuation.
KnowNow home screen in the safe state: a map around the user with the message “You are safe, no active hazards nearby.”
All clear
KnowNow home screen in the stay alert state: an orange hazard zone at the map edge, “You are safe,” and a button to see safe areas.
Stay alert
KnowNow home screen in the danger nearby state: a red hazard zone close to the user, a route line, and a “Start route to shelter” button.
Danger nearby
KnowNow home screen in the evacuate state: the user inside a dark red hazard zone and a full-width red “Evacuate now” panel.
Evacuate
A crisis close to home

The 2025 Los Angeles wildfires did not feel like a distant disaster. They affected the community around me, including classmates and faculty at ArtCenter. Beyond the visible destruction, I became interested in a less visible crisis: critical emergency information was not equally accessible across languages. Some non-English speakers could receive an alert yet still struggle to understand the danger, decide whether it affected them, or find help.

This led us to ask how emergency communication could become more accessible. But as the project developed, we discovered that translation was only one part of the gap between receiving an alert and knowing what to do next.

The people the system never sees

We interviewed multilingual volunteers, an interpreter, emergency responders, and people affected by the fires, and visited a Red Cross shelter. Inside, support appeared to work well. But that view only included people who had already understood the emergency and reached help. Others relied on family or community members to interpret alerts, while some felt uncomfortable asking for translation. The people facing the greatest language barriers could disappear before the formal support system ever saw them.

Field visit photos from a Red Cross emergency shelter in Los Angeles, January 2025.

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

Key insight
A system can appear effective while excluding the people who never make it through its first point of entry.
People scan, they don’t read.
Under pressure, long text and detailed explanations get skipped.
A/文
Understanding matters more than translation.
A translated alert can still be misread. Maps, icons, and color carry meaning faster than words.
“Am I safe?” comes first.
People need to see where they are relative to the risk before they can decide what to do.
Decisions happen in seconds.
The interface has to support immediate recognition, not exploration.
From universal access to personal action

By the end of the team project, we had expanded the challenge from helping non-English speakers seek support to creating universally accessible emergency communication. However, this framing was broader than the mobile experience we had designed. When I revisited the project independently, I narrowed the problem around a more immediate question: what does someone need to understand and do after receiving an alert?

Team’s final HMW
How might we design and implement universally accessible communication systems to overcome linguistic barriers during emergency response operations?
My reframed HMW
How might we help people understand how an emergency affects them and take the next safe action?
Core shift
From designing universal communication systems to guiding individual action during an emergency.
From parallel features to one sequence

The team concept offered five parallel tools: news, training, alerts, shelters, and a packing list. In the redesign, I rebuilt the system around a smaller set of core objects, each answering one question a person asks during an emergency.

1Alert
Am I in danger?
2Map
Where am I?
3Shelter
Where should I go?
4Route
How do I get there?

Each object exists to move people toward their next step. The system is no longer about what users can do, but about what they should do next.

Refined user flow: the user receives an alert, opens the map, the system checks whether they are in danger, then guides them to choose a shelter, start navigation, and follow the route to safety.

Refined user flow, from alert to a safe route

Testing the first screen

A notification is often the first contact with the app, so I tested the landing screen in two rounds of walkthroughs with an AI-simulated persona. These were not sessions with real users, but they showed where the screen failed to explain itself.

Round 1 landing screen: a map with a large red fire zone and a single button at the bottom.
Round 1
The red zone signaled danger, but not whether the user was inside it.
Round 2 landing screen: a “Danger nearby” label, the user’s position next to the red zone, a dotted route to a shelter, and a “Move to safety” panel with one primary button.
Round 2
Status, route, and one primary action on the first screen.
“I don’t know if I am in danger now. I don’t know if I should leave or stay.”

AI-simulated persona · Round 1 walkthrough

Say where the user stands

A status line such as “You are near the danger zone” or “You are inside the hazard zone” answers the first question before anything else.

Lead with one action

The bottom panel names the next step and offers a single primary button, instead of a generic “View details.”

Recommend one shelter

The nearest open shelter is highlighted rather than shown as one equal option among many.

One path from alert to action

I redesigned KnowNow as a map-based experience that connects each alert to the user’s location, current emergency state, available destinations, and next action.

KnowNow lock screen with three notifications, the newest reading “Danger! Fire 0.8 mi in NW. Evacuate now.”
KnowNow danger nearby screen: the user’s position beside a red hazard zone, a dotted route, and a “Move to safety” panel.
KnowNow safe shelters list sorted by distance, with the nearest shelter highlighted in green and capacity bars for each.
KnowNow shelter detail for Union Station Adult Center, showing capacity, available services, address, and a blue Navigate button.
KnowNow profile screen with live location sharing on, emergency contacts, medical and accessibility details, and language settings.
What happens
A lock-screen alert names the hazard, its distance, and the action to take.
Design decision
The alert leads with the action, so someone who only glances at the lock screen still knows what to do.
Supporting details
KnowNow map legend sheet listing disaster symbols: wildfire, flood, earthquake, and severe storm.
Map legend
Every zone, symbol, and shelter state is explained in one place.
KnowNow medical and accessibility screen listing conditions, allergies, and needs.
Medical & accessibility
Stored once, shared only when the user asks for help.
KnowNow home screen in dark mode, with an orange hazard zone and a “See safe areas” button.
Dark mode
The same states and colors, readable at night.

The redesign transformed KnowNow from a collection of emergency tools into a connected path from alert to action. Because the landing screen was refined through AI-simulated walkthroughs rather than sessions with real users, the next step would be to validate its language, emergency states, and action guidance with multilingual residents and emergency-response professionals.