Product Design · Wearable + App

Guardian Link

Role

Design lead on a team of four: user research, app interface, the pendant's 3D model and the logo. Teammates covered statistics, finance and mechanical engineering.

Duration

2023

Type

Team of four · UC Berkeley · Brief: Living Alone

Skills

User Research, Concept Selection, App UI, 3D Modeling

Tools

Figma, Fusion 360, 3D Printing

Coming home late, alone, in a city you just moved to. We designed a necklace that gets help without reaching for your phone.

Hold the pendant: your location goes to people you chose, and it lights up and sounds an alarm. The app shows what happened nearby.

A friend's alert
Near you tonight
Who gets the alert
Pendant status

The companion app, redesigned in 2026 from our 2023 concept.

Problem
The walk home at night is when people living alone feel least safe, and a phone is slow to reach.
What we learned
One button you can find without looking, worn by choice, sharing location on your terms.
What we made
A pendant that reads as jewelry, a hold-to-send alert, and an app that tells friends what to do next.

We started with living alone, not with safety

The brief was open: the barriers of living alone, from privacy and social life to health. We wrote 51 "how might we" questions and scored the strongest on six criteria.

Scored on

FeasibilityFit with our skillsShort-term marketLong-term marketBarriers to entryTeam interest

Safety came out on top: how can technology protect young people who live alone?

51"How might we" questions

Safety, top of the six criteria

Then we checked whether safety was worth solving

A high score doesn't mean it matters to people, so we went and asked.

15SurveyAll living alone, mostly 18 to 24.
2ndDeep-dive surveySecurity only.
NightParticipant observationWe walked the route home late at night.
7Target-user interviewsRated candidate features from 1 to 5.
3Competitor reviewA medical pendant and two alert watches.

Who we designed for

Matt, 19

First apartment after the dorm. Unsure the neighborhood is safe.

Jenny, 24

New job in California. Often walks home late through a high-crime area.

They wanted one button they could reach without thinking

6of 7

An SOS button came first.

Six gave it 5 out of 5. There's no time to unlock a phone.

Design principle

Reachable by touch, without a screen.

4of 7

It can't look like a medical alarm.

Four rated style highly. People want to wear it every day.

Design principle

Something you'd choose to wear.

4of 7

Location is sensitive.

Four rated privacy 4 out of 5. Users must control sharing.

Design principle

Nothing shared until you ask.

7of 7

Home is covered; the walk home isn't.

All seven already use home security like Ring. The gap is the street.

Design principle

Design for outside, not for home.

Competitors agreed: a medical pendant was reliable but nobody wanted to wear it. So we dropped home security and aimed for reliability people would choose to wear.

35 ideas, then a fair fight between the best six

From door cameras to wearables, narrowed in rounds.

  1. 35

    concepts

  2. 15

    after a first scoring

  3. 6

    directions sketched

  4. 1

    after two weighted matrices

35 concepts: how much tech, how far from what exists.

35 concepts: how much tech, how far from what exists.

Watch.

Watch.

Security cameras.

Security cameras.

The baseline changed the answer

Against a Ring doorbell, the smartwatch won. But watches already exist, so we reran it against an Apple Watch: what does each idea add beyond it?

Round 1 · baseline: Ring doorbell

1Smartwatch

2Necklace

Round 2 · baseline: Apple Watch

1Necklace

A necklace, because it is always on you

From our user-flow video: coming home alone at night.

From our user-flow video: coming home alone at night.

The only direction that answered all four findings.

Reachable without a screen

On your chest; find the button by touch.

Something you'd choose to wear

Designed to read as jewelry.

Nothing shared until you ask

Location goes only to your contacts, only when you press.

Outside, not home

Alerts go to your contact first, then the police.

Alerts that get attention without causing panic

Left: our 2023 screens. Right: my 2026 redesign.

A friend's alert · 2023

2023

A friend's alert · 2026

2026

01 · A friend's alert

Noticeable, not alarming.

The old alert just said "in danger". Now: who, where, how long ago, who else knows, and one next step: call them. 911 comes second.

Near you tonight · 2023

2023

Near you tonight · 2026

2026

02 · Near you tonight

Each report answers three questions.

What, how far, how long ago. Nothing to read on the walk home.

Who gets the alert · 2023

2023

Who gets the alert · 2026

2026

03 · Who gets the alert

Nothing is shared until you ask.

A page of toggles became two questions: who, and when. Night owls like Jenny can share automatically after dark.

Pendant status · 2023

2023

Pendant status · 2026

2026

04 · Pendant status

The pendant shows whether it is on you.

A closed cord means you're wearing it. Hold two seconds to send, so it won't go off by accident, and test it before you need it.

A pin inside a pin

You, inside the circle of people looking out for you. In 2026 I redrew it with even geometry and a cleaner typeface; the same blue runs through pendant and app.

Wordmark

Wordmark

App icon

App icon

The 2023 original

The 2023 original

From pipe cleaners to a printed pendant

Sketch model

01Sketch model

Pipe cleaners. How big, and where?

Working prototype

02Working prototype

A press-button lamp on a cord.

Enclosure

03Enclosure

Modeled in Fusion 360.

PLA print

04PLA print

Checked in the hand.

Next build: a microcontroller with Wi-Fi and Bluetooth, GPS, and an audio module for the alarm.

What I'd do differently

Fifteen surveys and seven users were enough to pick a direction, not to prove it. The 2026 features are hypotheses too. On a real walk home, I'd measure:

Seconds

to find and hold the button by touch

False alarms

per week

Days a week

people actually wear it

A risk we never discussed

An incident map can make people more afraid than the facts justify. A safety product should leave people calmer.