truepath: safety walking companion

Designing a quick action friendly looking app accessible style for the female solo commuters.

Project Overview

An AI-powered travel companion application that shifts the safety paradigm from reactive panic to proactive prevention by leveraging real-time environmental data. Subtle AI interactions will empower users to navigate nighttime cityscapes with confidence and peace of mind.

My responsibilities

truepath home screen on an iPhone

everyday commuting should not be worried about

encourage simple safe journey
in everyday life

truepath suggested routes screen on an iPhone

The Hard Question

How can we help women feel less unsafe when commuting alone.

Secondary research insights show

44%

of women feel unsafe when walking in the dark

50%

women felt unsafe walking alone after dark in a quiet street near their home.

compared to

men

1 in 7

What five women told us at midnight

"The anxiety peaks the moment they step off and walk alone"

Primary research insights from 5 interviewee show

Time Curfew

Phone call
Location Sharing

Environmental decay signals
unsafe areas (dark streets).

Step off the bus/tube to walk alone
makes them anxious

Gut-led threat detection

Women nearby are the strongest
safety proxy

Being foreign amplifies vulnerability
due to the unfamiliarity of the area

One-touch emergency
needed

Phone charge is a
non-negotiable safety
dependency

Reasons behind design decision: Designing for panic

A safety app that demands less of your attention. Your eyes should be on the street, not the screen.

Women don't need another app that reacts to emergencies.
They need one that prevents the emergency from happening.

Every feature was designed to work before danger arrives. Quiet data minimal interaction
and always one touch away from help.

AI Safest route map

Three data sources predict the safest path. Users see choices of route they can choose one they feel comfortable with.

Three data sources, one quiet prediction

With evidence-sourced, safe routes suggested are more trustworthy.

AI

crowd density

Bluetooth sensoring

OSM

Council API

Transit API

Deep Reinforcement
Learning

API

truepath on an iPhone

Haptic Navigation

Users said they navigate by gut feeling. Haptic vibrations guide without requiring them to look at their phone all the time.

eyes up

feel the way

Phone vibrating as navigation guides the user
Lock screen notification: turn left next crossroad

notifying navigation to users,
so they do not have to keep their phone unlocked.

Easy-touch SOS

None of women wants to use emergency access, but they still need it. Emergency access should be in under 3 seconds. One button, no unlocking, location shared automatically.

AI Fake call

The most common coping strategy women described was pretending to be on a phone call. I automated it into AI-integrated.

AI fake call screen 1 AI fake call screen 2 AI fake call screen 3
Easy-touch SOS screens

Design Approach

Dark mode

Because users might walk at night and a bright white screen kills their night vision and possibly makes them a visible target.

Blue branding

Allows the feeling of trust and calm without feeling cold or clinical.

Large touch targets

The user might be shaking or rushing, so emergency button should be easily touched at a glance.

Sans-serif typeface

Allows easily readable at a glance even in low light.

truepath logo
AI Friend
Call
Contact officer
Saved
Journey

SOS

Design system overview

What changed after testing with SUS scoring evaluation

3 out of 5 testing participants saying they might accidentally press SOS button during a simulated situation

I pull out a confirmation button to be easily pressed to prevent unintentional pressing.

SUS evaluation

truepath is built for everyday journey

users choose the AI Safe Route over the standard 'Fastest Route'  during a scenario test.
80%

What I’d do next